Skip to content
项目
群组
代码片段
帮助
当前项目
正在载入...
登录 / 注册
切换导航面板
T
Toy-MT-Introduction
概览
Overview
Details
Activity
Cycle Analytics
版本库
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
问题
0
Issues
0
列表
Board
标记
里程碑
合并请求
0
Merge Requests
0
CI / CD
CI / CD
流水线
作业
日程表
图表
维基
Wiki
代码片段
Snippets
成员
Collapse sidebar
Close sidebar
活动
图像
聊天
创建新问题
作业
提交
Issue Boards
Open sidebar
NiuTrans
Toy-MT-Introduction
Commits
623389ab
Commit
623389ab
authored
May 03, 2020
by
xiaotong
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
updates of the paper size and figures
parent
4aeb846b
显示空白字符变更
内嵌
并排
正在显示
5 个修改的文件
包含
516 行增加
和
491 行删除
+516
-491
Book/Chapter2/Figures/figure-analysis-of-sentence-participle&syntactic.tex
+1
-1
Book/Chapter7/Figures/progressive-training.tex
+4
-4
Book/mt-book-xelatex.idx
+145
-145
Book/mt-book-xelatex.ptc
+365
-341
Book/structure.tex
+1
-0
没有找到文件。
Book/Chapter2/Figures/figure-analysis-of-sentence-participle&syntactic.tex
查看文件 @
623389ab
...
@@ -6,7 +6,7 @@
...
@@ -6,7 +6,7 @@
\begin{tikzpicture}
\begin{tikzpicture}
\begin{scope}
[scale=1.0,
xshift=0.9in,yshift=-0.87in,level distance=20pt,sibling distance=-1
pt,grow'=up]
\begin{scope}
[scale=1.0,
level distance=30pt,sibling distance=15
pt,grow'=up]
{
{
\Tree
[.
\node
(sn0)
{
IP
}
;
\Tree
[.
\node
(sn0)
{
IP
}
;
[.
\node
(sn1)
{
NP
}
;
[.
\node
(sn1)
{
NP
}
;
...
...
Book/Chapter7/Figures/progressive-training.tex
查看文件 @
623389ab
...
@@ -22,10 +22,10 @@
...
@@ -22,10 +22,10 @@
\node
[anchor=west,fill=orange!20,draw=red,rounded corners=3pt,minimum height=1.4em,minimum width=1.4em,dashed] (s44) at ([xshift=1.5em]s43.east)
{$
\times
h
$}
;
\node
[anchor=west,fill=orange!20,draw=red,rounded corners=3pt,minimum height=1.4em,minimum width=1.4em,dashed] (s44) at ([xshift=1.5em]s43.east)
{$
\times
h
$}
;
\node
[anchor=west,fill=blue!20,draw=blue,rounded corners=3pt,minimum height=1.4em,minimum width=1.5em] (s45) at ([xshift=1.5em]s44.east)
{}
;
\node
[anchor=west,fill=blue!20,draw=blue,rounded corners=3pt,minimum height=1.4em,minimum width=1.5em] (s45) at ([xshift=1.5em]s44.east)
{}
;
\node
[anchor=east] (p1) at ([xshift=-2em]s11.west)
{
step1
}
;
\node
[anchor=east] (p1) at ([xshift=-2em]s11.west)
{
step
1
}
;
\node
[anchor=east] (p2) at ([xshift=-2em]s21.west)
{
step2
}
;
\node
[anchor=east] (p2) at ([xshift=-2em]s21.west)
{
step
2
}
;
\node
[anchor=east] (p3) at ([xshift=-2em]s31.west)
{
step3
}
;
\node
[anchor=east] (p3) at ([xshift=-2em]s31.west)
{
step
3
}
;
\node
[anchor=east] (p4) at ([xshift=-2em]s41.west)
{
step4
}
;
\node
[anchor=east] (p4) at ([xshift=-2em]s41.west)
{
step
4
}
;
\node
[anchor=south,fill=orange!20,draw=orange,rounded corners=3pt,minimum height=1.4em,minimum width=1.4em] (b1) at ([xshift=-0.2em,yshift=2em]p1.north)
{}
;
\node
[anchor=south,fill=orange!20,draw=orange,rounded corners=3pt,minimum height=1.4em,minimum width=1.4em] (b1) at ([xshift=-0.2em,yshift=2em]p1.north)
{}
;
\node
[anchor=west] (b2) at (b1.east)
{
：编码器
}
;
\node
[anchor=west] (b2) at (b1.east)
{
：编码器
}
;
...
...
Book/mt-book-xelatex.idx
查看文件 @
623389ab
\indexentry{未登录词|hyperpage}{1
1
}
\indexentry{未登录词|hyperpage}{1
7
}
\indexentry{Out of Vocabulary Word，OOV Word|hyperpage}{1
1
}
\indexentry{Out of Vocabulary Word，OOV Word|hyperpage}{1
7
}
\indexentry{子词切分|hyperpage}{1
1
}
\indexentry{子词切分|hyperpage}{1
7
}
\indexentry{Sub-word Segmentation|hyperpage}{1
1
}
\indexentry{Sub-word Segmentation|hyperpage}{1
7
}
\indexentry{标准化|hyperpage}{1
1
}
\indexentry{标准化|hyperpage}{1
7
}
\indexentry{Normalization|hyperpage}{1
1
}
\indexentry{Normalization|hyperpage}{1
7
}
\indexentry{数据清洗|hyperpage}{1
1
}
\indexentry{数据清洗|hyperpage}{1
7
}
\indexentry{Dada Cleaning|hyperpage}{1
1
}
\indexentry{Dada Cleaning|hyperpage}{1
7
}
\indexentry{数据选择|hyperpage}{1
3
}
\indexentry{数据选择|hyperpage}{1
9
}
\indexentry{Data Selection|hyperpage}{1
3
}
\indexentry{Data Selection|hyperpage}{1
9
}
\indexentry{数据过滤|hyperpage}{1
3
}
\indexentry{数据过滤|hyperpage}{1
9
}
\indexentry{Data Filtering|hyperpage}{1
3
}
\indexentry{Data Filtering|hyperpage}{1
9
}
\indexentry{开放词表|hyperpage}{
16
}
\indexentry{开放词表|hyperpage}{
22
}
\indexentry{Open-Vocabulary|hyperpage}{
16
}
\indexentry{Open-Vocabulary|hyperpage}{
22
}
\indexentry{子词|hyperpage}{
17
}
\indexentry{子词|hyperpage}{
23
}
\indexentry{Sub-word|hyperpage}{
17
}
\indexentry{Sub-word|hyperpage}{
23
}
\indexentry{字节对编码|hyperpage}{
17
}
\indexentry{字节对编码|hyperpage}{
23
}
\indexentry{双字节编码|hyperpage}{
17
}
\indexentry{双字节编码|hyperpage}{
23
}
\indexentry{Byte Pair Encoding，BPE|hyperpage}{
17
}
\indexentry{Byte Pair Encoding，BPE|hyperpage}{
23
}
\indexentry{正则化|hyperpage}{2
0
}
\indexentry{正则化|hyperpage}{2
6
}
\indexentry{Regularization|hyperpage}{2
0
}
\indexentry{Regularization|hyperpage}{2
6
}
\indexentry{过拟合问题|hyperpage}{2
0
}
\indexentry{过拟合问题|hyperpage}{2
6
}
\indexentry{Overfitting Problem|hyperpage}{2
0
}
\indexentry{Overfitting Problem|hyperpage}{2
6
}
\indexentry{反问题|hyperpage}{2
0
}
\indexentry{反问题|hyperpage}{2
6
}
\indexentry{Inverse Problem|hyperpage}{2
0
}
\indexentry{Inverse Problem|hyperpage}{2
6
}
\indexentry{适定的|hyperpage}{2
0
}
\indexentry{适定的|hyperpage}{2
6
}
\indexentry{Well-posed|hyperpage}{2
0
}
\indexentry{Well-posed|hyperpage}{2
6
}
\indexentry{不适定问题|hyperpage}{2
0
}
\indexentry{不适定问题|hyperpage}{2
6
}
\indexentry{Ill-posed Problem|hyperpage}{2
0
}
\indexentry{Ill-posed Problem|hyperpage}{2
6
}
\indexentry{降噪|hyperpage}{2
1
}
\indexentry{降噪|hyperpage}{2
7
}
\indexentry{Denoising|hyperpage}{2
1
}
\indexentry{Denoising|hyperpage}{2
7
}
\indexentry{泛化|hyperpage}{2
1
}
\indexentry{泛化|hyperpage}{2
7
}
\indexentry{Generalization|hyperpage}{2
1
}
\indexentry{Generalization|hyperpage}{2
7
}
\indexentry{标签平滑|hyperpage}{2
3
}
\indexentry{标签平滑|hyperpage}{2
9
}
\indexentry{Label Smoothing|hyperpage}{2
3
}
\indexentry{Label Smoothing|hyperpage}{2
9
}
\indexentry{相互适应|hyperpage}{
24
}
\indexentry{相互适应|hyperpage}{
30
}
\indexentry{Co-Adaptation|hyperpage}{
24
}
\indexentry{Co-Adaptation|hyperpage}{
30
}
\indexentry{集成学习|hyperpage}{
25
}
\indexentry{集成学习|hyperpage}{
31
}
\indexentry{Ensemble Learning|hyperpage}{
25
}
\indexentry{Ensemble Learning|hyperpage}{
31
}
\indexentry{容量|hyperpage}{
26
}
\indexentry{容量|hyperpage}{
32
}
\indexentry{Capacity|hyperpage}{
26
}
\indexentry{Capacity|hyperpage}{
32
}
\indexentry{宽残差网络|hyperpage}{
27
}
\indexentry{宽残差网络|hyperpage}{
33
}
\indexentry{Wide Residual Network|hyperpage}{
27
}
\indexentry{Wide Residual Network|hyperpage}{
33
}
\indexentry{探测任务|hyperpage}{
28
}
\indexentry{探测任务|hyperpage}{
34
}
\indexentry{Probing Task|hyperpage}{
28
}
\indexentry{Probing Task|hyperpage}{
34
}
\indexentry{表面信息|hyperpage}{
28
}
\indexentry{表面信息|hyperpage}{
34
}
\indexentry{Surface Information|hyperpage}{
28
}
\indexentry{Surface Information|hyperpage}{
34
}
\indexentry{语法信息|hyperpage}{
28
}
\indexentry{语法信息|hyperpage}{
34
}
\indexentry{Syntactic Information|hyperpage}{
28
}
\indexentry{Syntactic Information|hyperpage}{
34
}
\indexentry{语义信息|hyperpage}{
28
}
\indexentry{语义信息|hyperpage}{
34
}
\indexentry{Semantic Information|hyperpage}{
28
}
\indexentry{Semantic Information|hyperpage}{
34
}
\indexentry{词嵌入|hyperpage}{
29
}
\indexentry{词嵌入|hyperpage}{
35
}
\indexentry{Embedding|hyperpage}{
29
}
\indexentry{Embedding|hyperpage}{
35
}
\indexentry{数据并行|hyperpage}{
29
}
\indexentry{数据并行|hyperpage}{
35
}
\indexentry{Data Parallelism|hyperpage}{
29
}
\indexentry{Data Parallelism|hyperpage}{
35
}
\indexentry{模型并行|hyperpage}{
29
}
\indexentry{模型并行|hyperpage}{
35
}
\indexentry{Model Parallelism|hyperpage}{
29
}
\indexentry{Model Parallelism|hyperpage}{
35
}
\indexentry{小批量训练|hyperpage}{
29
}
\indexentry{小批量训练|hyperpage}{
35
}
\indexentry{Mini-batch Training|hyperpage}{
29
}
\indexentry{Mini-batch Training|hyperpage}{
35
}
\indexentry{课程学习|hyperpage}{3
1
}
\indexentry{课程学习|hyperpage}{3
7
}
\indexentry{Curriculum Learning|hyperpage}{3
1
}
\indexentry{Curriculum Learning|hyperpage}{3
7
}
\indexentry{推断|hyperpage}{3
2
}
\indexentry{推断|hyperpage}{3
8
}
\indexentry{Inference|hyperpage}{3
2
}
\indexentry{Inference|hyperpage}{3
8
}
\indexentry{解码|hyperpage}{3
2
}
\indexentry{解码|hyperpage}{3
8
}
\indexentry{Decoding|hyperpage}{3
2
}
\indexentry{Decoding|hyperpage}{3
8
}
\indexentry{搜索错误|hyperpage}{3
2
}
\indexentry{搜索错误|hyperpage}{3
8
}
\indexentry{Search Error|hyperpage}{3
2
}
\indexentry{Search Error|hyperpage}{3
8
}
\indexentry{模型错误|hyperpage}{3
2
}
\indexentry{模型错误|hyperpage}{3
8
}
\indexentry{Modeling Error|hyperpage}{3
2
}
\indexentry{Modeling Error|hyperpage}{3
8
}
\indexentry{重排序|hyperpage}{
34
}
\indexentry{重排序|hyperpage}{
40
}
\indexentry{Re-ranking|hyperpage}{
34
}
\indexentry{Re-ranking|hyperpage}{
40
}
\indexentry{双向推断|hyperpage}{
34
}
\indexentry{双向推断|hyperpage}{
40
}
\indexentry{Bidirectional Inference|hyperpage}{
34
}
\indexentry{Bidirectional Inference|hyperpage}{
40
}
\indexentry{批量推断|hyperpage}{
38
}
\indexentry{批量推断|hyperpage}{
44
}
\indexentry{Batch Inference|hyperpage}{
38
}
\indexentry{Batch Inference|hyperpage}{
44
}
\indexentry{批量处理|hyperpage}{
38
}
\indexentry{批量处理|hyperpage}{
44
}
\indexentry{Batching|hyperpage}{
38
}
\indexentry{Batching|hyperpage}{
44
}
\indexentry{二值网络|hyperpage}{
39
}
\indexentry{二值网络|hyperpage}{
45
}
\indexentry{Binarized Neural Networks|hyperpage}{
39
}
\indexentry{Binarized Neural Networks|hyperpage}{
45
}
\indexentry{自回归翻译|hyperpage}{4
0
}
\indexentry{自回归翻译|hyperpage}{4
6
}
\indexentry{Autoregressive Translation|hyperpage}{4
0
}
\indexentry{Autoregressive Translation|hyperpage}{4
6
}
\indexentry{非自回归翻译|hyperpage}{4
0
}
\indexentry{非自回归翻译|hyperpage}{4
6
}
\indexentry{Regressive Translation|hyperpage}{4
0
}
\indexentry{Regressive Translation|hyperpage}{4
6
}
\indexentry{繁衍率|hyperpage}{4
0
}
\indexentry{繁衍率|hyperpage}{4
6
}
\indexentry{Fertility|hyperpage}{4
0
}
\indexentry{Fertility|hyperpage}{4
6
}
\indexentry{偏置|hyperpage}{4
1
}
\indexentry{偏置|hyperpage}{4
7
}
\indexentry{Bias|hyperpage}{4
1
}
\indexentry{Bias|hyperpage}{4
7
}
\indexentry{退化|hyperpage}{4
2
}
\indexentry{退化|hyperpage}{4
8
}
\indexentry{Degenerate|hyperpage}{4
2
}
\indexentry{Degenerate|hyperpage}{4
8
}
\indexentry{过翻译|hyperpage}{4
3
}
\indexentry{过翻译|hyperpage}{4
9
}
\indexentry{Over Translation|hyperpage}{4
3
}
\indexentry{Over Translation|hyperpage}{4
9
}
\indexentry{欠翻译|hyperpage}{4
3
}
\indexentry{欠翻译|hyperpage}{4
9
}
\indexentry{Under Translation|hyperpage}{4
3
}
\indexentry{Under Translation|hyperpage}{4
9
}
\indexentry{充分性|hyperpage}{
44
}
\indexentry{充分性|hyperpage}{
50
}
\indexentry{Adequacy|hyperpage}{
44
}
\indexentry{Adequacy|hyperpage}{
50
}
\indexentry{系统融合|hyperpage}{
44
}
\indexentry{系统融合|hyperpage}{
50
}
\indexentry{System Combination|hyperpage}{
44
}
\indexentry{System Combination|hyperpage}{
50
}
\indexentry{假设选择|hyperpage}{
45
}
\indexentry{假设选择|hyperpage}{
51
}
\indexentry{Hypothesis Selection|hyperpage}{
45
}
\indexentry{Hypothesis Selection|hyperpage}{
51
}
\indexentry{多样性|hyperpage}{
45
}
\indexentry{多样性|hyperpage}{
51
}
\indexentry{Diversity|hyperpage}{
45
}
\indexentry{Diversity|hyperpage}{
51
}
\indexentry{重排序|hyperpage}{
46
}
\indexentry{重排序|hyperpage}{
52
}
\indexentry{Re-ranking|hyperpage}{
46
}
\indexentry{Re-ranking|hyperpage}{
52
}
\indexentry{混淆网络|hyperpage}{
47
}
\indexentry{混淆网络|hyperpage}{
53
}
\indexentry{Confusion Network|hyperpage}{
47
}
\indexentry{Confusion Network|hyperpage}{
53
}
\indexentry{动态线性层聚合方法|hyperpage}{5
1
}
\indexentry{动态线性层聚合方法|hyperpage}{5
7
}
\indexentry{Dynamic Linear Combination of Layers，DLCL|hyperpage}{5
1
}
\indexentry{Dynamic Linear Combination of Layers，DLCL|hyperpage}{5
7
}
\indexentry{相互适应|hyperpage}{
55
}
\indexentry{相互适应|hyperpage}{
61
}
\indexentry{Co-adaptation|hyperpage}{
55
}
\indexentry{Co-adaptation|hyperpage}{
61
}
\indexentry{数据增强|hyperpage}{
57
}
\indexentry{数据增强|hyperpage}{
63
}
\indexentry{Data Augmentation|hyperpage}{
57
}
\indexentry{Data Augmentation|hyperpage}{
63
}
\indexentry{回译|hyperpage}{
57
}
\indexentry{回译|hyperpage}{
63
}
\indexentry{Back Translation|hyperpage}{
57
}
\indexentry{Back Translation|hyperpage}{
63
}
\indexentry{迭代式回译|hyperpage}{
58
}
\indexentry{迭代式回译|hyperpage}{
64
}
\indexentry{Iterative Back Translation|hyperpage}{
58
}
\indexentry{Iterative Back Translation|hyperpage}{
64
}
\indexentry{前向翻译|hyperpage}{
58
}
\indexentry{前向翻译|hyperpage}{
64
}
\indexentry{Forward Translation|hyperpage}{
58
}
\indexentry{Forward Translation|hyperpage}{
64
}
\indexentry{预训练|hyperpage}{
59
}
\indexentry{预训练|hyperpage}{
65
}
\indexentry{Pre-training|hyperpage}{
59
}
\indexentry{Pre-training|hyperpage}{
65
}
\indexentry{微调|hyperpage}{
59
}
\indexentry{微调|hyperpage}{
65
}
\indexentry{Fine-tuning|hyperpage}{
59
}
\indexentry{Fine-tuning|hyperpage}{
65
}
\indexentry{多任务学习|hyperpage}{6
1
}
\indexentry{多任务学习|hyperpage}{6
7
}
\indexentry{Multitask Learning|hyperpage}{6
1
}
\indexentry{Multitask Learning|hyperpage}{6
7
}
\indexentry{模型压缩|hyperpage}{6
2
}
\indexentry{模型压缩|hyperpage}{6
8
}
\indexentry{Model Compression|hyperpage}{6
2
}
\indexentry{Model Compression|hyperpage}{6
8
}
\indexentry{学习难度|hyperpage}{6
2
}
\indexentry{学习难度|hyperpage}{6
8
}
\indexentry{Learning Difficulty|hyperpage}{6
2
}
\indexentry{Learning Difficulty|hyperpage}{6
8
}
\indexentry{教师模型|hyperpage}{6
3
}
\indexentry{教师模型|hyperpage}{6
9
}
\indexentry{Teacher Model|hyperpage}{6
3
}
\indexentry{Teacher Model|hyperpage}{6
9
}
\indexentry{学生模型|hyperpage}{6
3
}
\indexentry{学生模型|hyperpage}{6
9
}
\indexentry{Student Model|hyperpage}{6
3
}
\indexentry{Student Model|hyperpage}{6
9
}
\indexentry{基于单词的知识精炼|hyperpage}{6
3
}
\indexentry{基于单词的知识精炼|hyperpage}{6
9
}
\indexentry{Word-level Knowledge Distillation|hyperpage}{6
3
}
\indexentry{Word-level Knowledge Distillation|hyperpage}{6
9
}
\indexentry{基于序列的知识精炼|hyperpage}{6
3
}
\indexentry{基于序列的知识精炼|hyperpage}{6
9
}
\indexentry{Sequence-level Knowledge Distillation|hyperpage}{6
3
}
\indexentry{Sequence-level Knowledge Distillation|hyperpage}{6
9
}
\indexentry{中间层输出|hyperpage}{
64
}
\indexentry{中间层输出|hyperpage}{
70
}
\indexentry{Hint-based Knowledge Transfer|hyperpage}{
64
}
\indexentry{Hint-based Knowledge Transfer|hyperpage}{
70
}
\indexentry{注意力分布|hyperpage}{
64
}
\indexentry{注意力分布|hyperpage}{
70
}
\indexentry{Attention To Attention Transfer|hyperpage}{
64
}
\indexentry{Attention To Attention Transfer|hyperpage}{
70
}
\indexentry{循环一致性|hyperpage}{
67
}
\indexentry{循环一致性|hyperpage}{
73
}
\indexentry{Circle Consistency|hyperpage}{
67
}
\indexentry{Circle Consistency|hyperpage}{
73
}
\indexentry{翻译中回译|hyperpage}{
68
}
\indexentry{翻译中回译|hyperpage}{
74
}
\indexentry{On-the-fly Back-translation|hyperpage}{
68
}
\indexentry{On-the-fly Back-translation|hyperpage}{
74
}
\indexentry{网络结构搜索技术|hyperpage}{7
1
}
\indexentry{网络结构搜索技术|hyperpage}{7
7
}
\indexentry{Neural Architecture Search；NAS|hyperpage}{7
1
}
\indexentry{Neural Architecture Search；NAS|hyperpage}{7
7
}
Book/mt-book-xelatex.ptc
查看文件 @
623389ab
...
@@ -2,690 +2,714 @@
...
@@ -2,690 +2,714 @@
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\select@language {english}
\select@language {english}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {part}{\@mypartnumtocformat {I}{机器翻译基础}}{1
1
}{part.1}
\contentsline {part}{\@mypartnumtocformat {I}{机器翻译基础}}{1
3
}{part.1}
\ttl@starttoc {default@1}
\ttl@starttoc {default@1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {1}机器翻译简介}{1
3
}{chapter.1}
\contentsline {chapter}{\numberline {1}机器翻译简介}{1
5
}{chapter.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.1}机器翻译的概念}{1
3
}{section.1.1}
\contentsline {section}{\numberline {1.1}机器翻译的概念}{1
5
}{section.1.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.2}机器翻译简史}{1
6
}{section.1.2}
\contentsline {section}{\numberline {1.2}机器翻译简史}{1
8
}{section.1.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.2.1}人工翻译}{1
6
}{subsection.1.2.1}
\contentsline {subsection}{\numberline {1.2.1}人工翻译}{1
8
}{subsection.1.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.2.2}机器翻译的萌芽}{1
7
}{subsection.1.2.2}
\contentsline {subsection}{\numberline {1.2.2}机器翻译的萌芽}{1
9
}{subsection.1.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.2.3}机器翻译的受挫}{
18
}{subsection.1.2.3}
\contentsline {subsection}{\numberline {1.2.3}机器翻译的受挫}{
20
}{subsection.1.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.2.4}机器翻译的快速成长}{
19
}{subsection.1.2.4}
\contentsline {subsection}{\numberline {1.2.4}机器翻译的快速成长}{
21
}{subsection.1.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.2.5}机器翻译的爆发}{2
0
}{subsection.1.2.5}
\contentsline {subsection}{\numberline {1.2.5}机器翻译的爆发}{2
2
}{subsection.1.2.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.3}机器翻译现状}{2
1
}{section.1.3}
\contentsline {section}{\numberline {1.3}机器翻译现状}{2
3
}{section.1.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.4}机器翻译方法}{2
2
}{section.1.4}
\contentsline {section}{\numberline {1.4}机器翻译方法}{2
4
}{section.1.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.4.1}基于规则的机器翻译}{2
4
}{subsection.1.4.1}
\contentsline {subsection}{\numberline {1.4.1}基于规则的机器翻译}{2
6
}{subsection.1.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.4.2}基于实例的机器翻译}{2
4
}{subsection.1.4.2}
\contentsline {subsection}{\numberline {1.4.2}基于实例的机器翻译}{2
6
}{subsection.1.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.4.3}统计机器翻译}{2
5
}{subsection.1.4.3}
\contentsline {subsection}{\numberline {1.4.3}统计机器翻译}{2
7
}{subsection.1.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.4.4}神经机器翻译}{2
6
}{subsection.1.4.4}
\contentsline {subsection}{\numberline {1.4.4}神经机器翻译}{2
8
}{subsection.1.4.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.4.5}对比分析}{2
7
}{subsection.1.4.5}
\contentsline {subsection}{\numberline {1.4.5}对比分析}{2
9
}{subsection.1.4.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.5}翻译质量评价}{
28
}{section.1.5}
\contentsline {section}{\numberline {1.5}翻译质量评价}{
30
}{section.1.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.5.1}人工评价}{
28
}{subsection.1.5.1}
\contentsline {subsection}{\numberline {1.5.1}人工评价}{
30
}{subsection.1.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.5.2}自动评价}{
29
}{subsection.1.5.2}
\contentsline {subsection}{\numberline {1.5.2}自动评价}{
31
}{subsection.1.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{BLEU}{
29
}{section*.15}
\contentsline {subsubsection}{BLEU}{
31
}{section*.15}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{TER}{3
1
}{section*.16}
\contentsline {subsubsection}{TER}{3
3
}{section*.16}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于检测点的评价}{3
1
}{section*.17}
\contentsline {subsubsection}{基于检测点的评价}{3
3
}{section*.17}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.6}机器翻译应用}{3
2
}{section.1.6}
\contentsline {section}{\numberline {1.6}机器翻译应用}{3
4
}{section.1.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.7}开源项目与评测}{3
4
}{section.1.7}
\contentsline {section}{\numberline {1.7}开源项目与评测}{3
6
}{section.1.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.7.1}开源机器翻译系统}{3
4
}{subsection.1.7.1}
\contentsline {subsection}{\numberline {1.7.1}开源机器翻译系统}{3
6
}{subsection.1.7.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{统计机器翻译开源系统}{3
5
}{section*.19}
\contentsline {subsubsection}{统计机器翻译开源系统}{3
7
}{section*.19}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{神经机器翻译开源系统}{3
6
}{section*.20}
\contentsline {subsubsection}{神经机器翻译开源系统}{3
8
}{section*.20}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {1.7.2}常用数据集及公开评测任务}{
38
}{subsection.1.7.2}
\contentsline {subsection}{\numberline {1.7.2}常用数据集及公开评测任务}{
40
}{subsection.1.7.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {1.8}推荐学习资源}{4
0
}{section.1.8}
\contentsline {section}{\numberline {1.8}推荐学习资源}{4
2
}{section.1.8}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {2}词法、语法及统计建模基础}{4
5
}{chapter.2}
\contentsline {chapter}{\numberline {2}词法、语法及统计建模基础}{4
7
}{chapter.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {2.1}问题概述 }{4
6
}{section.2.1}
\contentsline {section}{\numberline {2.1}问题概述 }{4
8
}{section.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {2.2}概率论基础}{4
7
}{section.2.2}
\contentsline {section}{\numberline {2.2}概率论基础}{4
9
}{section.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.2.1}随机变量和概率}{
47
}{subsection.2.2.1}
\contentsline {subsection}{\numberline {2.2.1}随机变量和概率}{
50
}{subsection.2.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.2.2}联合概率、条件概率和边缘概率}{
49
}{subsection.2.2.2}
\contentsline {subsection}{\numberline {2.2.2}联合概率、条件概率和边缘概率}{
51
}{subsection.2.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.2.3}链式法则}{5
0
}{subsection.2.2.3}
\contentsline {subsection}{\numberline {2.2.3}链式法则}{5
2
}{subsection.2.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.2.4}贝叶斯法则}{5
1
}{subsection.2.2.4}
\contentsline {subsection}{\numberline {2.2.4}贝叶斯法则}{5
3
}{subsection.2.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.2.5}KL距离和熵}{5
3
}{subsection.2.2.5}
\contentsline {subsection}{\numberline {2.2.5}KL距离和熵}{5
5
}{subsection.2.2.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{信息熵}{5
3
}{section*.27}
\contentsline {subsubsection}{信息熵}{5
5
}{section*.27}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{KL距离}{5
4
}{section*.29}
\contentsline {subsubsection}{KL距离}{5
6
}{section*.29}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{交叉熵}{5
4
}{section*.30}
\contentsline {subsubsection}{交叉熵}{5
6
}{section*.30}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {2.3}中文分词}{5
5
}{section.2.3}
\contentsline {section}{\numberline {2.3}中文分词}{5
7
}{section.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.3.1}基于词典的分词方法}{5
6
}{subsection.2.3.1}
\contentsline {subsection}{\numberline {2.3.1}基于词典的分词方法}{5
8
}{subsection.2.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.3.2}基于统计的分词方法}{5
7
}{subsection.2.3.2}
\contentsline {subsection}{\numberline {2.3.2}基于统计的分词方法}{5
9
}{subsection.2.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{统计模型的学习与推断}{5
7
}{section*.34}
\contentsline {subsubsection}{统计模型的学习与推断}{5
9
}{section*.34}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{掷骰子游戏}{
58
}{section*.36}
\contentsline {subsubsection}{掷骰子游戏}{
60
}{section*.36}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{全概率分词方法}{6
0
}{section*.40}
\contentsline {subsubsection}{全概率分词方法}{6
2
}{section*.40}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {2.4}$n$-gram语言模型 }{6
2
}{section.2.4}
\contentsline {section}{\numberline {2.4}$n$-gram语言模型 }{6
4
}{section.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.4.1}建模}{6
3
}{subsection.2.4.1}
\contentsline {subsection}{\numberline {2.4.1}建模}{6
5
}{subsection.2.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.4.2}未登录词和平滑算法}{6
5
}{subsection.2.4.2}
\contentsline {subsection}{\numberline {2.4.2}未登录词和平滑算法}{6
7
}{subsection.2.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{加法平滑方法}{6
6
}{section*.46}
\contentsline {subsubsection}{加法平滑方法}{6
8
}{section*.46}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{古德-图灵估计法}{6
7
}{section*.48}
\contentsline {subsubsection}{古德-图灵估计法}{6
9
}{section*.48}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{Kneser-Ney平滑方法}{
68
}{section*.50}
\contentsline {subsubsection}{Kneser-Ney平滑方法}{
70
}{section*.50}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {2.5}句法分析（短语结构分析）}{7
0
}{section.2.5}
\contentsline {section}{\numberline {2.5}句法分析（短语结构分析）}{7
2
}{section.2.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.5.1}句子的句法树表示}{7
0
}{subsection.2.5.1}
\contentsline {subsection}{\numberline {2.5.1}句子的句法树表示}{7
2
}{subsection.2.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.5.2}上下文无关文法}{7
2
}{subsection.2.5.2}
\contentsline {subsection}{\numberline {2.5.2}上下文无关文法}{7
4
}{subsection.2.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {2.5.3}规则和推导的概率}{7
6
}{subsection.2.5.3}
\contentsline {subsection}{\numberline {2.5.3}规则和推导的概率}{7
8
}{subsection.2.5.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {2.6}小结及深入阅读}{
78
}{section.2.6}
\contentsline {section}{\numberline {2.6}小结及深入阅读}{
80
}{section.2.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {part}{\@mypartnumtocformat {II}{统计机器翻译}}{8
1
}{part.2}
\contentsline {part}{\@mypartnumtocformat {II}{统计机器翻译}}{8
3
}{part.2}
\ttl@stoptoc {default@1}
\ttl@stoptoc {default@1}
\ttl@starttoc {default@2}
\ttl@starttoc {default@2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {3}基于词的机器翻译模型}{8
3
}{chapter.3}
\contentsline {chapter}{\numberline {3}基于词的机器翻译模型}{8
5
}{chapter.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.1}什么是基于词的翻译模型}{8
3
}{section.3.1}
\contentsline {section}{\numberline {3.1}什么是基于词的翻译模型}{8
5
}{section.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.2}构建一个简单的机器翻译系统}{8
5
}{section.3.2}
\contentsline {section}{\numberline {3.2}构建一个简单的机器翻译系统}{8
7
}{section.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.2.1}如何进行翻译？}{8
5
}{subsection.3.2.1}
\contentsline {subsection}{\numberline {3.2.1}如何进行翻译？}{8
7
}{subsection.3.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{机器翻译流程}{8
6
}{section*.63}
\contentsline {subsubsection}{机器翻译流程}{8
8
}{section*.63}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{人工翻译 vs. 机器翻译}{8
7
}{section*.65}
\contentsline {subsubsection}{人工翻译 vs. 机器翻译}{8
9
}{section*.65}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.2.2}基本框架}{8
7
}{subsection.3.2.2}
\contentsline {subsection}{\numberline {3.2.2}基本框架}{8
9
}{subsection.3.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.2.3}单词翻译概率}{
88
}{subsection.3.2.3}
\contentsline {subsection}{\numberline {3.2.3}单词翻译概率}{
90
}{subsection.3.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{什么是单词翻译概率？}{
88
}{section*.67}
\contentsline {subsubsection}{什么是单词翻译概率？}{
90
}{section*.67}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{如何从一个双语平行数据中学习？}{
88
}{section*.69}
\contentsline {subsubsection}{如何从一个双语平行数据中学习？}{
90
}{section*.69}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{如何从大量的双语平行数据中学习？}{9
0
}{section*.70}
\contentsline {subsubsection}{如何从大量的双语平行数据中学习？}{9
2
}{section*.70}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.2.4}句子级翻译模型}{9
1
}{subsection.3.2.4}
\contentsline {subsection}{\numberline {3.2.4}句子级翻译模型}{9
3
}{subsection.3.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基础模型}{9
1
}{section*.72}
\contentsline {subsubsection}{基础模型}{9
3
}{section*.72}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{生成流畅的译文}{9
3
}{section*.74}
\contentsline {subsubsection}{生成流畅的译文}{9
5
}{section*.74}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.2.5}解码}{9
5
}{subsection.3.2.5}
\contentsline {subsection}{\numberline {3.2.5}解码}{9
7
}{subsection.3.2.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.3}基于词的翻译建模}{
98
}{section.3.3}
\contentsline {section}{\numberline {3.3}基于词的翻译建模}{
100
}{section.3.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.3.1}噪声信道模型}{
98
}{subsection.3.3.1}
\contentsline {subsection}{\numberline {3.3.1}噪声信道模型}{
100
}{subsection.3.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.3.2}统计机器翻译的三个基本问题}{10
0
}{subsection.3.3.2}
\contentsline {subsection}{\numberline {3.3.2}统计机器翻译的三个基本问题}{10
2
}{subsection.3.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{词对齐}{10
1
}{section*.83}
\contentsline {subsubsection}{词对齐}{10
3
}{section*.83}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于词对齐的翻译模型}{10
1
}{section*.86}
\contentsline {subsubsection}{基于词对齐的翻译模型}{10
3
}{section*.86}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于词对齐的翻译实例}{10
3
}{section*.88}
\contentsline {subsubsection}{基于词对齐的翻译实例}{10
5
}{section*.88}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.4}IBM模型1-2}{10
4
}{section.3.4}
\contentsline {section}{\numberline {3.4}IBM模型1-2}{10
6
}{section.3.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.4.1}IBM模型1}{10
4
}{subsection.3.4.1}
\contentsline {subsection}{\numberline {3.4.1}IBM模型1}{10
6
}{subsection.3.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.4.2}IBM模型2}{10
6
}{subsection.3.4.2}
\contentsline {subsection}{\numberline {3.4.2}IBM模型2}{10
8
}{subsection.3.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.4.3}解码及计算优化}{10
7
}{subsection.3.4.3}
\contentsline {subsection}{\numberline {3.4.3}解码及计算优化}{10
9
}{subsection.3.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.4.4}训练}{1
08
}{subsection.3.4.4}
\contentsline {subsection}{\numberline {3.4.4}训练}{1
10
}{subsection.3.4.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{目标函数}{1
08
}{section*.93}
\contentsline {subsubsection}{目标函数}{1
10
}{section*.93}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{优化}{1
09
}{section*.95}
\contentsline {subsubsection}{优化}{1
11
}{section*.95}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.5}IBM模型3-5及隐马尔可夫模型}{11
5
}{section.3.5}
\contentsline {section}{\numberline {3.5}IBM模型3-5及隐马尔可夫模型}{11
7
}{section.3.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.5.1}基于产出率的翻译模型}{11
5
}{subsection.3.5.1}
\contentsline {subsection}{\numberline {3.5.1}基于产出率的翻译模型}{11
7
}{subsection.3.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.5.2}IBM 模型3}{1
18
}{subsection.3.5.2}
\contentsline {subsection}{\numberline {3.5.2}IBM 模型3}{1
20
}{subsection.3.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.5.3}IBM 模型4}{1
19
}{subsection.3.5.3}
\contentsline {subsection}{\numberline {3.5.3}IBM 模型4}{1
21
}{subsection.3.5.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.5.4} IBM 模型5}{12
1
}{subsection.3.5.4}
\contentsline {subsection}{\numberline {3.5.4} IBM 模型5}{12
3
}{subsection.3.5.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.5.5}隐马尔可夫模型}{12
2
}{subsection.3.5.5}
\contentsline {subsection}{\numberline {3.5.5}隐马尔可夫模型}{12
4
}{subsection.3.5.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{隐马尔可夫模型}{12
3
}{section*.107}
\contentsline {subsubsection}{隐马尔可夫模型}{12
5
}{section*.107}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{词对齐模型}{12
4
}{section*.109}
\contentsline {subsubsection}{词对齐模型}{12
6
}{section*.109}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.5.6}解码和训练}{12
5
}{subsection.3.5.6}
\contentsline {subsection}{\numberline {3.5.6}解码和训练}{12
7
}{subsection.3.5.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.6}问题分析}{12
5
}{section.3.6}
\contentsline {section}{\numberline {3.6}问题分析}{12
7
}{section.3.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.6.1}词对齐及对称化}{12
5
}{subsection.3.6.1}
\contentsline {subsection}{\numberline {3.6.1}词对齐及对称化}{12
7
}{subsection.3.6.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.6.2}Deficiency}{12
6
}{subsection.3.6.2}
\contentsline {subsection}{\numberline {3.6.2}Deficiency}{12
8
}{subsection.3.6.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.6.3}句子长度}{12
7
}{subsection.3.6.3}
\contentsline {subsection}{\numberline {3.6.3}句子长度}{12
9
}{subsection.3.6.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {3.6.4}其他问题}{1
28
}{subsection.3.6.4}
\contentsline {subsection}{\numberline {3.6.4}其他问题}{1
30
}{subsection.3.6.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {3.7}小结及深入阅读}{1
28
}{section.3.7}
\contentsline {section}{\numberline {3.7}小结及深入阅读}{1
30
}{section.3.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {4}基于短语和句法的机器翻译模型}{13
1
}{chapter.4}
\contentsline {chapter}{\numberline {4}基于短语和句法的机器翻译模型}{13
3
}{chapter.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {4.1}翻译中的结构信息}{13
1
}{section.4.1}
\contentsline {section}{\numberline {4.1}翻译中的结构信息}{13
3
}{section.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.1.1}更大粒度的翻译单元}{13
2
}{subsection.4.1.1}
\contentsline {subsection}{\numberline {4.1.1}更大粒度的翻译单元}{13
4
}{subsection.4.1.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.1.2}句子的结构信息}{13
4
}{subsection.4.1.2}
\contentsline {subsection}{\numberline {4.1.2}句子的结构信息}{13
6
}{subsection.4.1.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {4.2}基于短语的翻译模型}{13
6
}{section.4.2}
\contentsline {section}{\numberline {4.2}基于短语的翻译模型}{13
8
}{section.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.1}机器翻译中的短语}{13
6
}{subsection.4.2.1}
\contentsline {subsection}{\numberline {4.2.1}机器翻译中的短语}{13
8
}{subsection.4.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.2}数学建模及判别式模型}{1
39
}{subsection.4.2.2}
\contentsline {subsection}{\numberline {4.2.2}数学建模及判别式模型}{1
41
}{subsection.4.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于翻译推导的建模}{1
39
}{section*.121}
\contentsline {subsubsection}{基于翻译推导的建模}{1
41
}{section*.121}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{对数线性模型}{14
0
}{section*.122}
\contentsline {subsubsection}{对数线性模型}{14
2
}{section*.122}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{搭建模型的基本流程}{14
1
}{section*.123}
\contentsline {subsubsection}{搭建模型的基本流程}{14
3
}{section*.123}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.3}短语抽取}{14
2
}{subsection.4.2.3}
\contentsline {subsection}{\numberline {4.2.3}短语抽取}{14
4
}{subsection.4.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{与词对齐一致的短语}{14
3
}{section*.126}
\contentsline {subsubsection}{与词对齐一致的短语}{14
5
}{section*.126}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{获取词对齐}{14
4
}{section*.130}
\contentsline {subsubsection}{获取词对齐}{14
6
}{section*.130}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{度量双语短语质量}{14
5
}{section*.132}
\contentsline {subsubsection}{度量双语短语质量}{14
7
}{section*.132}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.4}调序}{14
6
}{subsection.4.2.4}
\contentsline {subsection}{\numberline {4.2.4}调序}{14
8
}{subsection.4.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于距离的调序}{14
6
}{section*.136}
\contentsline {subsubsection}{基于距离的调序}{14
8
}{section*.136}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于方向的调序}{14
7
}{section*.138}
\contentsline {subsubsection}{基于方向的调序}{14
9
}{section*.138}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于分类的调序}{1
49
}{section*.141}
\contentsline {subsubsection}{基于分类的调序}{1
51
}{section*.141}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.5}特征}{1
49
}{subsection.4.2.5}
\contentsline {subsection}{\numberline {4.2.5}特征}{1
51
}{subsection.4.2.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.6}最小错误率训练}{15
0
}{subsection.4.2.6}
\contentsline {subsection}{\numberline {4.2.6}最小错误率训练}{15
2
}{subsection.4.2.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.2.7}栈解码}{15
3
}{subsection.4.2.7}
\contentsline {subsection}{\numberline {4.2.7}栈解码}{15
5
}{subsection.4.2.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{翻译候选匹配}{15
4
}{section*.146}
\contentsline {subsubsection}{翻译候选匹配}{15
6
}{section*.146}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{翻译假设扩展}{15
4
}{section*.148}
\contentsline {subsubsection}{翻译假设扩展}{15
6
}{section*.148}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{剪枝}{15
5
}{section*.150}
\contentsline {subsubsection}{剪枝}{15
7
}{section*.150}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{解码中的栈结构}{15
7
}{section*.152}
\contentsline {subsubsection}{解码中的栈结构}{15
9
}{section*.152}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {4.3}基于层次短语的模型}{1
58
}{section.4.3}
\contentsline {section}{\numberline {4.3}基于层次短语的模型}{1
60
}{section.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.3.1}同步上下文无关文法}{16
1
}{subsection.4.3.1}
\contentsline {subsection}{\numberline {4.3.1}同步上下文无关文法}{16
3
}{subsection.4.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{文法定义}{16
1
}{section*.157}
\contentsline {subsubsection}{文法定义}{16
3
}{section*.157}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{推导}{16
2
}{section*.158}
\contentsline {subsubsection}{推导}{16
4
}{section*.158}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{胶水规则}{16
3
}{section*.159}
\contentsline {subsubsection}{胶水规则}{16
5
}{section*.159}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{处理流程}{16
4
}{section*.160}
\contentsline {subsubsection}{处理流程}{16
6
}{section*.160}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.3.2}层次短语规则抽取}{16
4
}{subsection.4.3.2}
\contentsline {subsection}{\numberline {4.3.2}层次短语规则抽取}{16
6
}{subsection.4.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.3.3}翻译模型及特征}{16
6
}{subsection.4.3.3}
\contentsline {subsection}{\numberline {4.3.3}翻译模型及特征}{16
8
}{subsection.4.3.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.3.4}CYK解码}{16
7
}{subsection.4.3.4}
\contentsline {subsection}{\numberline {4.3.4}CYK解码}{16
9
}{subsection.4.3.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.3.5}立方剪枝}{17
0
}{subsection.4.3.5}
\contentsline {subsection}{\numberline {4.3.5}立方剪枝}{17
2
}{subsection.4.3.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {4.4}基于语言学句法的模型}{17
3
}{section.4.4}
\contentsline {section}{\numberline {4.4}基于语言学句法的模型}{17
5
}{section.4.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.1}基于句法的翻译模型分类}{17
5
}{subsection.4.4.1}
\contentsline {subsection}{\numberline {4.4.1}基于句法的翻译模型分类}{17
7
}{subsection.4.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.2}基于树结构的文法}{17
5
}{subsection.4.4.2}
\contentsline {subsection}{\numberline {4.4.2}基于树结构的文法}{17
7
}{subsection.4.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{树到树翻译规则}{17
7
}{section*.176}
\contentsline {subsubsection}{树到树翻译规则}{17
9
}{section*.176}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于树结构的翻译推导}{1
79
}{section*.178}
\contentsline {subsubsection}{基于树结构的翻译推导}{1
81
}{section*.178}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{树到串翻译规则}{18
1
}{section*.181}
\contentsline {subsubsection}{树到串翻译规则}{18
3
}{section*.181}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.3}树到串翻译规则抽取}{18
2
}{subsection.4.4.3}
\contentsline {subsection}{\numberline {4.4.3}树到串翻译规则抽取}{18
4
}{subsection.4.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{树的切割与最小规则}{18
3
}{section*.183}
\contentsline {subsubsection}{树的切割与最小规则}{18
5
}{section*.183}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{空对齐处理}{18
6
}{section*.189}
\contentsline {subsubsection}{空对齐处理}{18
8
}{section*.189}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{组合规则}{18
7
}{section*.191}
\contentsline {subsubsection}{组合规则}{18
9
}{section*.191}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{SPMT规则}{1
88
}{section*.193}
\contentsline {subsubsection}{SPMT规则}{1
90
}{section*.193}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{句法树二叉化}{1
89
}{section*.195}
\contentsline {subsubsection}{句法树二叉化}{1
91
}{section*.195}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.4}树到树翻译规则抽取}{19
0
}{subsection.4.4.4}
\contentsline {subsection}{\numberline {4.4.4}树到树翻译规则抽取}{19
2
}{subsection.4.4.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于节点对齐的规则抽取}{19
1
}{section*.199}
\contentsline {subsubsection}{基于节点对齐的规则抽取}{19
3
}{section*.199}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于对齐矩阵的规则抽取}{19
2
}{section*.202}
\contentsline {subsubsection}{基于对齐矩阵的规则抽取}{19
4
}{section*.202}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.5}句法翻译模型的特征}{19
4
}{subsection.4.4.5}
\contentsline {subsection}{\numberline {4.4.5}句法翻译模型的特征}{19
6
}{subsection.4.4.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.6}基于超图的推导空间表示}{19
5
}{subsection.4.4.6}
\contentsline {subsection}{\numberline {4.4.6}基于超图的推导空间表示}{19
7
}{subsection.4.4.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {4.4.7}基于树的解码 vs 基于串的解码}{19
7
}{subsection.4.4.7}
\contentsline {subsection}{\numberline {4.4.7}基于树的解码 vs 基于串的解码}{19
9
}{subsection.4.4.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于树的解码}{
199
}{section*.209}
\contentsline {subsubsection}{基于树的解码}{
201
}{section*.209}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于串的解码}{20
0
}{section*.212}
\contentsline {subsubsection}{基于串的解码}{20
2
}{section*.212}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {4.5}小结及深入阅读}{20
2
}{section.4.5}
\contentsline {section}{\numberline {4.5}小结及深入阅读}{20
4
}{section.4.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {part}{\@mypartnumtocformat {III}{神经机器翻译}}{20
5
}{part.3}
\contentsline {part}{\@mypartnumtocformat {III}{神经机器翻译}}{20
7
}{part.3}
\ttl@stoptoc {default@2}
\ttl@stoptoc {default@2}
\ttl@starttoc {default@3}
\ttl@starttoc {default@3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {5}人工神经网络和神经语言建模}{20
7
}{chapter.5}
\contentsline {chapter}{\numberline {5}人工神经网络和神经语言建模}{20
9
}{chapter.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {5.1}深度学习与人工神经网络}{2
08
}{section.5.1}
\contentsline {section}{\numberline {5.1}深度学习与人工神经网络}{2
10
}{section.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.1.1}发展简史}{2
08
}{subsection.5.1.1}
\contentsline {subsection}{\numberline {5.1.1}发展简史}{2
10
}{subsection.5.1.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{早期的人工神经网络和第一次寒冬}{2
08
}{section*.214}
\contentsline {subsubsection}{早期的人工神经网络和第一次寒冬}{2
10
}{section*.214}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{神经网络的第二次高潮和第二次寒冬}{2
09
}{section*.215}
\contentsline {subsubsection}{神经网络的第二次高潮和第二次寒冬}{2
11
}{section*.215}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{深度学习和神经网络方法的崛起}{21
0
}{section*.216}
\contentsline {subsubsection}{深度学习和神经网络方法的崛起}{21
2
}{section*.216}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.1.2}为什么需要深度学习}{21
1
}{subsection.5.1.2}
\contentsline {subsection}{\numberline {5.1.2}为什么需要深度学习}{21
3
}{subsection.5.1.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{端到端学习和表示学习}{21
1
}{section*.218}
\contentsline {subsubsection}{端到端学习和表示学习}{21
3
}{section*.218}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{深度学习的效果}{21
2
}{section*.220}
\contentsline {subsubsection}{深度学习的效果}{21
4
}{section*.220}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {5.2}神经网络基础}{21
2
}{section.5.2}
\contentsline {section}{\numberline {5.2}神经网络基础}{21
4
}{section.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.2.1}线性代数基础}{21
2
}{subsection.5.2.1}
\contentsline {subsection}{\numberline {5.2.1}线性代数基础}{21
4
}{subsection.5.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{标量、向量和矩阵}{21
3
}{section*.222}
\contentsline {subsubsection}{标量、向量和矩阵}{21
5
}{section*.222}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{矩阵的转置}{21
4
}{section*.223}
\contentsline {subsubsection}{矩阵的转置}{21
6
}{section*.223}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{矩阵加法和数乘}{21
4
}{section*.224}
\contentsline {subsubsection}{矩阵加法和数乘}{21
6
}{section*.224}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{矩阵乘法和矩阵点乘}{21
5
}{section*.225}
\contentsline {subsubsection}{矩阵乘法和矩阵点乘}{21
7
}{section*.225}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{线性映射}{21
6
}{section*.226}
\contentsline {subsubsection}{线性映射}{21
8
}{section*.226}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{范数}{21
7
}{section*.227}
\contentsline {subsubsection}{范数}{21
9
}{section*.227}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.2.2}人工神经元和感知机}{2
18
}{subsection.5.2.2}
\contentsline {subsection}{\numberline {5.2.2}人工神经元和感知机}{2
20
}{subsection.5.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{感知机\ \raisebox {0.5mm}{------}\ 最简单的人工神经元模型}{2
19
}{section*.230}
\contentsline {subsubsection}{感知机\ \raisebox {0.5mm}{------}\ 最简单的人工神经元模型}{2
21
}{section*.230}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{神经元内部权重}{22
0
}{section*.233}
\contentsline {subsubsection}{神经元内部权重}{22
2
}{section*.233}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{神经元的输入\ \raisebox {0.5mm}{------}\ 离散 vs 连续}{22
1
}{section*.235}
\contentsline {subsubsection}{神经元的输入\ \raisebox {0.5mm}{------}\ 离散 vs 连续}{22
3
}{section*.235}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{神经元内部的参数学习}{22
1
}{section*.237}
\contentsline {subsubsection}{神经元内部的参数学习}{22
3
}{section*.237}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.2.3}多层神经网络}{22
2
}{subsection.5.2.3}
\contentsline {subsection}{\numberline {5.2.3}多层神经网络}{22
4
}{subsection.5.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{线性变换和激活函数}{22
2
}{section*.239}
\contentsline {subsubsection}{线性变换和激活函数}{22
4
}{section*.239}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{单层神经网络$\rightarrow $多层神经网络}{22
4
}{section*.246}
\contentsline {subsubsection}{单层神经网络$\rightarrow $多层神经网络}{22
6
}{section*.246}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.2.4}函数拟合能力}{22
5
}{subsection.5.2.4}
\contentsline {subsection}{\numberline {5.2.4}函数拟合能力}{22
7
}{subsection.5.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {5.3}神经网络的张量实现}{2
29
}{section.5.3}
\contentsline {section}{\numberline {5.3}神经网络的张量实现}{2
31
}{section.5.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.3.1} 张量及其计算}{23
0
}{subsection.5.3.1}
\contentsline {subsection}{\numberline {5.3.1} 张量及其计算}{23
2
}{subsection.5.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{张量}{23
0
}{section*.256}
\contentsline {subsubsection}{张量}{23
2
}{section*.256}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{张量的矩阵乘法}{23
2
}{section*.259}
\contentsline {subsubsection}{张量的矩阵乘法}{23
4
}{section*.259}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{张量的单元操作}{23
3
}{section*.261}
\contentsline {subsubsection}{张量的单元操作}{23
5
}{section*.261}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.3.2}张量的物理存储形式}{23
4
}{subsection.5.3.2}
\contentsline {subsection}{\numberline {5.3.2}张量的物理存储形式}{23
6
}{subsection.5.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.3.3}使用开源框架实现张量计算}{23
4
}{subsection.5.3.3}
\contentsline {subsection}{\numberline {5.3.3}使用开源框架实现张量计算}{23
6
}{subsection.5.3.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.3.4}前向传播与计算图}{23
6
}{subsection.5.3.4}
\contentsline {subsection}{\numberline {5.3.4}前向传播与计算图}{23
8
}{subsection.5.3.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.3.5}神经网络实例}{2
39
}{subsection.5.3.5}
\contentsline {subsection}{\numberline {5.3.5}神经网络实例}{2
41
}{subsection.5.3.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {5.4}神经网络的参数训练}{24
0
}{section.5.4}
\contentsline {section}{\numberline {5.4}神经网络的参数训练}{24
2
}{section.5.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.4.1}损失函数}{24
1
}{subsection.5.4.1}
\contentsline {subsection}{\numberline {5.4.1}损失函数}{24
3
}{subsection.5.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.4.2}基于梯度的参数优化}{24
1
}{subsection.5.4.2}
\contentsline {subsection}{\numberline {5.4.2}基于梯度的参数优化}{24
3
}{subsection.5.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{梯度下降}{24
2
}{section*.279}
\contentsline {subsubsection}{梯度下降}{24
4
}{section*.279}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{梯度获取}{24
4
}{section*.281}
\contentsline {subsubsection}{梯度获取}{24
6
}{section*.281}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于梯度的方法的变种和改进}{24
7
}{section*.285}
\contentsline {subsubsection}{基于梯度的方法的变种和改进}{24
9
}{section*.285}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.4.3}参数更新的并行化策略}{25
0
}{subsection.5.4.3}
\contentsline {subsection}{\numberline {5.4.3}参数更新的并行化策略}{25
2
}{subsection.5.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.4.4}梯度消失、梯度爆炸和稳定性训练}{25
2
}{subsection.5.4.4}
\contentsline {subsection}{\numberline {5.4.4}梯度消失、梯度爆炸和稳定性训练}{25
4
}{subsection.5.4.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{易于优化的激活函数}{25
2
}{section*.288}
\contentsline {subsubsection}{易于优化的激活函数}{25
4
}{section*.288}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{梯度裁剪}{25
3
}{section*.292}
\contentsline {subsubsection}{梯度裁剪}{25
5
}{section*.292}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{稳定性训练}{25
4
}{section*.293}
\contentsline {subsubsection}{稳定性训练}{25
6
}{section*.293}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.4.5}过拟合}{25
5
}{subsection.5.4.5}
\contentsline {subsection}{\numberline {5.4.5}过拟合}{25
7
}{subsection.5.4.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.4.6}反向传播}{25
6
}{subsection.5.4.6}
\contentsline {subsection}{\numberline {5.4.6}反向传播}{25
8
}{subsection.5.4.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{输出层的反向传播}{25
7
}{section*.296}
\contentsline {subsubsection}{输出层的反向传播}{25
9
}{section*.296}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{隐藏层的反向传播}{2
59
}{section*.300}
\contentsline {subsubsection}{隐藏层的反向传播}{2
61
}{section*.300}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{程序实现}{26
0
}{section*.303}
\contentsline {subsubsection}{程序实现}{26
2
}{section*.303}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {5.5}神经语言模型}{26
2
}{section.5.5}
\contentsline {section}{\numberline {5.5}神经语言模型}{26
4
}{section.5.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.5.1}基于神经网络的语言建模}{26
2
}{subsection.5.5.1}
\contentsline {subsection}{\numberline {5.5.1}基于神经网络的语言建模}{26
4
}{subsection.5.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于前馈神经网络的语言模型}{26
3
}{section*.306}
\contentsline {subsubsection}{基于前馈神经网络的语言模型}{26
5
}{section*.306}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于循环神经网络的语言模型}{26
5
}{section*.309}
\contentsline {subsubsection}{基于循环神经网络的语言模型}{26
7
}{section*.309}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{基于自注意力机制的语言模型}{26
6
}{section*.311}
\contentsline {subsubsection}{基于自注意力机制的语言模型}{26
8
}{section*.311}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{语言模型的评价}{26
7
}{section*.313}
\contentsline {subsubsection}{语言模型的评价}{26
9
}{section*.313}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.5.2}单词表示模型}{2
68
}{subsection.5.5.2}
\contentsline {subsection}{\numberline {5.5.2}单词表示模型}{2
70
}{subsection.5.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{One-hot编码}{2
68
}{section*.314}
\contentsline {subsubsection}{One-hot编码}{2
70
}{section*.314}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{分布式表示}{2
68
}{section*.316}
\contentsline {subsubsection}{分布式表示}{2
70
}{section*.316}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {5.5.3}句子表示模型及预训练}{27
0
}{subsection.5.5.3}
\contentsline {subsection}{\numberline {5.5.3}句子表示模型及预训练}{27
2
}{subsection.5.5.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{简单的上下文表示模型}{27
0
}{section*.320}
\contentsline {subsubsection}{简单的上下文表示模型}{27
2
}{section*.320}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{ELMO模型}{27
2
}{section*.323}
\contentsline {subsubsection}{ELMO模型}{27
4
}{section*.323}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{GPT模型}{27
2
}{section*.325}
\contentsline {subsubsection}{GPT模型}{27
4
}{section*.325}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{BERT模型}{27
3
}{section*.327}
\contentsline {subsubsection}{BERT模型}{27
5
}{section*.327}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{为什么要预训练？}{27
4
}{section*.329}
\contentsline {subsubsection}{为什么要预训练？}{27
6
}{section*.329}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {5.6}小结及深入阅读}{27
5
}{section.5.6}
\contentsline {section}{\numberline {5.6}小结及深入阅读}{27
7
}{section.5.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {6}神经机器翻译模型}{27
7
}{chapter.6}
\contentsline {chapter}{\numberline {6}神经机器翻译模型}{27
9
}{chapter.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {6.1}神经机器翻译的发展简史}{27
7
}{section.6.1}
\contentsline {section}{\numberline {6.1}神经机器翻译的发展简史}{27
9
}{section.6.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.1.1}神经机器翻译的起源}{2
79
}{subsection.6.1.1}
\contentsline {subsection}{\numberline {6.1.1}神经机器翻译的起源}{2
81
}{subsection.6.1.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.1.2}神经机器翻译的品质 }{28
1
}{subsection.6.1.2}
\contentsline {subsection}{\numberline {6.1.2}神经机器翻译的品质 }{28
3
}{subsection.6.1.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.1.3}神经机器翻译的优势 }{28
4
}{subsection.6.1.3}
\contentsline {subsection}{\numberline {6.1.3}神经机器翻译的优势 }{28
6
}{subsection.6.1.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {6.2}编码器-解码器框架}{28
6
}{section.6.2}
\contentsline {section}{\numberline {6.2}编码器-解码器框架}{28
8
}{section.6.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.2.1}框架结构}{28
6
}{subsection.6.2.1}
\contentsline {subsection}{\numberline {6.2.1}框架结构}{28
8
}{subsection.6.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.2.2}表示学习}{28
7
}{subsection.6.2.2}
\contentsline {subsection}{\numberline {6.2.2}表示学习}{28
9
}{subsection.6.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.2.3}简单的运行实例}{2
88
}{subsection.6.2.3}
\contentsline {subsection}{\numberline {6.2.3}简单的运行实例}{2
90
}{subsection.6.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.2.4}机器翻译范式的对比}{2
89
}{subsection.6.2.4}
\contentsline {subsection}{\numberline {6.2.4}机器翻译范式的对比}{2
91
}{subsection.6.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {6.3}基于循环神经网络的翻译模型及注意力机制}{29
0
}{section.6.3}
\contentsline {section}{\numberline {6.3}基于循环神经网络的翻译模型及注意力机制}{29
2
}{section.6.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.1}建模}{29
0
}{subsection.6.3.1}
\contentsline {subsection}{\numberline {6.3.1}建模}{29
2
}{subsection.6.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.2}输入（词嵌入）及输出（Softmax）}{29
4
}{subsection.6.3.2}
\contentsline {subsection}{\numberline {6.3.2}输入（词嵌入）及输出（Softmax）}{29
6
}{subsection.6.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.3}循环神经网络结构}{
298
}{subsection.6.3.3}
\contentsline {subsection}{\numberline {6.3.3}循环神经网络结构}{
300
}{subsection.6.3.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{循环神经单元（RNN）}{
298
}{section*.351}
\contentsline {subsubsection}{循环神经单元（RNN）}{
300
}{section*.351}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{长短时记忆网络（LSTM）}{
298
}{section*.352}
\contentsline {subsubsection}{长短时记忆网络（LSTM）}{
300
}{section*.352}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{门控循环单元（GRU）}{30
0
}{section*.355}
\contentsline {subsubsection}{门控循环单元（GRU）}{30
2
}{section*.355}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{双向模型}{30
2
}{section*.357}
\contentsline {subsubsection}{双向模型}{30
4
}{section*.357}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{多层循环神经网络}{30
2
}{section*.359}
\contentsline {subsubsection}{多层循环神经网络}{30
4
}{section*.359}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.4}注意力机制}{30
3
}{subsection.6.3.4}
\contentsline {subsection}{\numberline {6.3.4}注意力机制}{30
5
}{subsection.6.3.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{翻译中的注意力机制}{30
4
}{section*.362}
\contentsline {subsubsection}{翻译中的注意力机制}{30
6
}{section*.362}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{上下文向量的计算}{30
5
}{section*.365}
\contentsline {subsubsection}{上下文向量的计算}{30
7
}{section*.365}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{注意力机制的解读}{3
08
}{section*.370}
\contentsline {subsubsection}{注意力机制的解读}{3
10
}{section*.370}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.5}训练}{31
0
}{subsection.6.3.5}
\contentsline {subsection}{\numberline {6.3.5}训练}{31
2
}{subsection.6.3.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{损失函数}{31
0
}{section*.373}
\contentsline {subsubsection}{损失函数}{31
2
}{section*.373}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{长参数初始化}{31
1
}{section*.374}
\contentsline {subsubsection}{长参数初始化}{31
3
}{section*.374}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{优化策略}{31
2
}{section*.375}
\contentsline {subsubsection}{优化策略}{31
4
}{section*.375}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{梯度裁剪}{31
2
}{section*.377}
\contentsline {subsubsection}{梯度裁剪}{31
4
}{section*.377}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{学习率策略}{31
2
}{section*.378}
\contentsline {subsubsection}{学习率策略}{31
4
}{section*.378}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{并行训练}{31
4
}{section*.381}
\contentsline {subsubsection}{并行训练}{31
6
}{section*.381}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.6}推断}{31
5
}{subsection.6.3.6}
\contentsline {subsection}{\numberline {6.3.6}推断}{31
7
}{subsection.6.3.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{贪婪搜索}{31
7
}{section*.385}
\contentsline {subsubsection}{贪婪搜索}{31
9
}{section*.385}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{束搜索}{3
18
}{section*.388}
\contentsline {subsubsection}{束搜索}{3
20
}{section*.388}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{长度惩罚}{3
19
}{section*.390}
\contentsline {subsubsection}{长度惩罚}{3
21
}{section*.390}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.3.7}实例-GNMT}{32
0
}{subsection.6.3.7}
\contentsline {subsection}{\numberline {6.3.7}实例-GNMT}{32
2
}{subsection.6.3.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {6.4}Transformer}{32
1
}{section.6.4}
\contentsline {section}{\numberline {6.4}Transformer}{32
3
}{section.6.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.1}自注意力模型}{32
3
}{subsection.6.4.1}
\contentsline {subsection}{\numberline {6.4.1}自注意力模型}{32
5
}{subsection.6.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.2}Transformer架构}{32
4
}{subsection.6.4.2}
\contentsline {subsection}{\numberline {6.4.2}Transformer架构}{32
6
}{subsection.6.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.3}位置编码}{32
6
}{subsection.6.4.3}
\contentsline {subsection}{\numberline {6.4.3}位置编码}{32
8
}{subsection.6.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.4}基于点乘的注意力机制}{3
29
}{subsection.6.4.4}
\contentsline {subsection}{\numberline {6.4.4}基于点乘的注意力机制}{3
31
}{subsection.6.4.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.5}掩码操作}{33
1
}{subsection.6.4.5}
\contentsline {subsection}{\numberline {6.4.5}掩码操作}{33
3
}{subsection.6.4.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.6}多头注意力}{33
2
}{subsection.6.4.6}
\contentsline {subsection}{\numberline {6.4.6}多头注意力}{33
4
}{subsection.6.4.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.7}残差网络和层正则化}{33
3
}{subsection.6.4.7}
\contentsline {subsection}{\numberline {6.4.7}残差网络和层正则化}{33
5
}{subsection.6.4.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.8}前馈全连接网络子层}{33
4
}{subsection.6.4.8}
\contentsline {subsection}{\numberline {6.4.8}前馈全连接网络子层}{33
6
}{subsection.6.4.8}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.9}训练}{33
5
}{subsection.6.4.9}
\contentsline {subsection}{\numberline {6.4.9}训练}{33
7
}{subsection.6.4.9}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.4.10}推断}{3
38
}{subsection.6.4.10}
\contentsline {subsection}{\numberline {6.4.10}推断}{3
40
}{subsection.6.4.10}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {6.5}序列到序列问题及应用}{3
38
}{section.6.5}
\contentsline {section}{\numberline {6.5}序列到序列问题及应用}{3
40
}{section.6.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.5.1}自动问答}{3
39
}{subsection.6.5.1}
\contentsline {subsection}{\numberline {6.5.1}自动问答}{3
41
}{subsection.6.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.5.2}自动文摘}{3
39
}{subsection.6.5.2}
\contentsline {subsection}{\numberline {6.5.2}自动文摘}{3
41
}{subsection.6.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.5.3}文言文翻译}{34
0
}{subsection.6.5.3}
\contentsline {subsection}{\numberline {6.5.3}文言文翻译}{34
2
}{subsection.6.5.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.5.4}对联生成}{34
0
}{subsection.6.5.4}
\contentsline {subsection}{\numberline {6.5.4}对联生成}{34
2
}{subsection.6.5.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {6.5.5}古诗生成}{34
1
}{subsection.6.5.5}
\contentsline {subsection}{\numberline {6.5.5}古诗生成}{34
3
}{subsection.6.5.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {6.6}小结及深入阅读}{34
2
}{section.6.6}
\contentsline {section}{\numberline {6.6}小结及深入阅读}{34
4
}{section.6.6}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {7}神经机器翻译实战 \ \raisebox {0.5mm}{------}\ 参加一次比赛}{34
5
}{chapter.7}
\contentsline {chapter}{\numberline {7}神经机器翻译实战 \ \raisebox {0.5mm}{------}\ 参加一次比赛}{34
7
}{chapter.7}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {7.1}神经机器翻译并不简单}{34
5
}{section.7.1}
\contentsline {section}{\numberline {7.1}神经机器翻译并不简单}{34
7
}{section.7.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.1.1}影响神经机器翻译性能的因素}{34
6
}{subsection.7.1.1}
\contentsline {subsection}{\numberline {7.1.1}影响神经机器翻译性能的因素}{34
8
}{subsection.7.1.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.1.2}搭建神经机器翻译系统的步骤 }{34
7
}{subsection.7.1.2}
\contentsline {subsection}{\numberline {7.1.2}搭建神经机器翻译系统的步骤 }{34
9
}{subsection.7.1.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.1.3}架构选择 }{3
48
}{subsection.7.1.3}
\contentsline {subsection}{\numberline {7.1.3}架构选择 }{3
50
}{subsection.7.1.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {7.2}数据处理}{3
48
}{section.7.2}
\contentsline {section}{\numberline {7.2}数据处理}{3
50
}{section.7.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.2.1}分词}{3
49
}{subsection.7.2.1}
\contentsline {subsection}{\numberline {7.2.1}分词}{3
51
}{subsection.7.2.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.2.2}标准化}{35
0
}{subsection.7.2.2}
\contentsline {subsection}{\numberline {7.2.2}标准化}{35
2
}{subsection.7.2.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.2.3}数据清洗}{35
1
}{subsection.7.2.3}
\contentsline {subsection}{\numberline {7.2.3}数据清洗}{35
3
}{subsection.7.2.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.2.4}子词切分}{35
3
}{subsection.7.2.4}
\contentsline {subsection}{\numberline {7.2.4}子词切分}{35
5
}{subsection.7.2.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{大词表和OOV问题}{35
4
}{section*.428}
\contentsline {subsubsection}{大词表和OOV问题}{35
6
}{section*.428}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{子词}{35
4
}{section*.430}
\contentsline {subsubsection}{子词}{35
6
}{section*.430}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{双字节编码（BPE）}{35
5
}{section*.432}
\contentsline {subsubsection}{双字节编码（BPE）}{35
7
}{section*.432}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{其他方法}{3
58
}{section*.435}
\contentsline {subsubsection}{其他方法}{3
60
}{section*.435}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {7.3}建模与训练}{3
58
}{section.7.3}
\contentsline {section}{\numberline {7.3}建模与训练}{3
60
}{section.7.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.3.1}正则化}{3
58
}{subsection.7.3.1}
\contentsline {subsection}{\numberline {7.3.1}正则化}{3
60
}{subsection.7.3.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{L1/L2正则化}{36
0
}{section*.437}
\contentsline {subsubsection}{L1/L2正则化}{36
2
}{section*.437}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{标签平滑}{36
1
}{section*.438}
\contentsline {subsubsection}{标签平滑}{36
3
}{section*.438}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{Dropout}{36
1
}{section*.440}
\contentsline {subsubsection}{Dropout}{36
4
}{section*.440}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{Layer Dropout}{36
3
}{section*.443}
\contentsline {subsubsection}{Layer Dropout}{36
5
}{section*.443}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.3.2}增大模型容量}{36
4
}{subsection.7.3.2}
\contentsline {subsection}{\numberline {7.3.2}增大模型容量}{36
6
}{subsection.7.3.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{宽网络}{36
4
}{section*.445}
\contentsline {subsubsection}{宽网络}{36
6
}{section*.445}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{深网络}{36
5
}{section*.447}
\contentsline {subsubsection}{深网络}{36
7
}{section*.447}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{增大输入层和输出层表示能力}{36
6
}{section*.449}
\contentsline {subsubsection}{增大输入层和输出层表示能力}{36
9
}{section*.449}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{大模型的分布式计算}{36
7
}{section*.450}
\contentsline {subsubsection}{大模型的分布式计算}{36
9
}{section*.450}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.3.3}大批量训练}{36
7
}{subsection.7.3.3}
\contentsline {subsection}{\numberline {7.3.3}大批量训练}{36
9
}{subsection.7.3.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{为什么需要大批量训练}{3
67
}{section*.451}
\contentsline {subsubsection}{为什么需要大批量训练}{3
70
}{section*.451}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{如何构建批次}{3
69
}{section*.454}
\contentsline {subsubsection}{如何构建批次}{3
71
}{section*.454}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {7.4}推断}{37
0
}{section.7.4}
\contentsline {section}{\numberline {7.4}推断}{37
2
}{section.7.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.4.1}推断优化}{37
0
}{subsection.7.4.1}
\contentsline {subsection}{\numberline {7.4.1}推断优化}{37
2
}{subsection.7.4.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{推断系统的架构}{37
0
}{section*.456}
\contentsline {subsubsection}{推断系统的架构}{37
2
}{section*.456}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{自左向右推断 vs 自右向左推断}{37
1
}{section*.458}
\contentsline {subsubsection}{自左向右推断 vs 自右向左推断}{37
3
}{section*.458}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{推断加速}{37
2
}{section*.459}
\contentsline {subsubsection}{推断加速}{37
4
}{section*.459}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.4.2}译文长度控制}{3
79
}{subsection.7.4.2}
\contentsline {subsection}{\numberline {7.4.2}译文长度控制}{3
81
}{subsection.7.4.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{长度惩罚因子}{3
79
}{section*.465}
\contentsline {subsubsection}{长度惩罚因子}{3
82
}{section*.465}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{译文长度范围约束}{38
0
}{section*.467}
\contentsline {subsubsection}{译文长度范围约束}{38
3
}{section*.467}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{覆盖度模型}{38
1
}{section*.468}
\contentsline {subsubsection}{覆盖度模型}{38
3
}{section*.468}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.4.3}多模型集成}{38
2
}{subsection.7.4.3}
\contentsline {subsection}{\numberline {7.4.3}多模型集成}{38
4
}{subsection.7.4.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{假设选择}{38
2
}{section*.469}
\contentsline {subsubsection}{假设选择}{38
5
}{section*.469}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{局部预测融合}{38
3
}{section*.471}
\contentsline {subsubsection}{局部预测融合}{38
6
}{section*.471}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{译文重组}{38
4
}{section*.473}
\contentsline {subsubsection}{译文重组}{38
7
}{section*.473}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {7.5}进阶技术}{38
5
}{section.7.5}
\contentsline {section}{\numberline {7.5}进阶技术}{38
8
}{section.7.5}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.5.1}深层模型}{38
5
}{subsection.7.5.1}
\contentsline {subsection}{\numberline {7.5.1}深层模型}{38
8
}{subsection.7.5.1}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{Post-Norm vs Pre-Norm}{38
6
}{section*.476}
\contentsline {subsubsection}{Post-Norm vs Pre-Norm}{38
8
}{section*.476}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{层聚合}{3
88
}{section*.479}
\contentsline {subsubsection}{层聚合}{3
91
}{section*.479}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{深层模型的训练加速}{3
89
}{section*.481}
\contentsline {subsubsection}{深层模型的训练加速}{3
92
}{section*.481}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{渐进式训练}{39
0
}{section*.482}
\contentsline {subsubsection}{渐进式训练}{39
2
}{section*.482}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{分组稠密连接}{39
0
}{section*.484}
\contentsline {subsubsection}{分组稠密连接}{39
2
}{section*.484}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{学习率重置策略}{39
1
}{section*.486}
\contentsline {subsubsection}{学习率重置策略}{39
3
}{section*.486}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{深层模型的鲁棒性训练}{39
2
}{section*.488}
\contentsline {subsubsection}{深层模型的鲁棒性训练}{39
5
}{section*.488}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.5.2}单语数据的使用}{39
4
}{subsection.7.5.2}
\contentsline {subsection}{\numberline {7.5.2}单语数据的使用}{39
6
}{subsection.7.5.2}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{伪数据}{39
5
}{section*.491}
\contentsline {subsubsection}{伪数据}{39
7
}{section*.491}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{预训练}{39
6
}{section*.494}
\contentsline {subsubsection}{预训练}{39
9
}{section*.494}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{联合训练}{
398
}{section*.497}
\contentsline {subsubsection}{联合训练}{
401
}{section*.497}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.5.3}知识精炼}{
399
}{subsection.7.5.3}
\contentsline {subsection}{\numberline {7.5.3}知识精炼}{
401
}{subsection.7.5.3}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{什么是知识精炼}{
399
}{section*.499}
\contentsline {subsubsection}{什么是知识精炼}{
402
}{section*.499}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{知识精炼的基本方法}{40
1
}{section*.500}
\contentsline {subsubsection}{知识精炼的基本方法}{40
3
}{section*.500}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{机器翻译中的知识精炼}{40
2
}{section*.502}
\contentsline {subsubsection}{机器翻译中的知识精炼}{40
4
}{section*.502}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {7.5.4}双向训练}{40
3
}{subsection.7.5.4}
\contentsline {subsection}{\numberline {7.5.4}双向训练}{40
6
}{subsection.7.5.4}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{有监督对偶学习}{40
4
}{section*.504}
\contentsline {subsubsection}{有监督对偶学习}{40
6
}{section*.504}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{无监督对偶学习}{40
5
}{section*.505}
\contentsline {subsubsection}{无监督对偶学习}{40
7
}{section*.505}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {subsubsection}{翻译中回译}{40
6
}{section*.507}
\contentsline {subsubsection}{翻译中回译}{40
8
}{section*.507}
\defcounter {refsection}{0}\relax
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {7.6}小结及深入阅读}{406}{section.7.6}
\contentsline {section}{\numberline {7.6}小结及深入阅读}{408}{section.7.6}
\defcounter {refsection}{0}\relax
\contentsline {part}{\@mypartnumtocformat {IV}{附录}}{413}{part.4}
\ttl@stoptoc {default@3}
\ttl@starttoc {default@4}
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {A}附录A}{415}{Appendix.1.A}
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {A.1}基准数据集}{415}{section.1.A.1}
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {A.2}平行语料}{416}{section.1.A.2}
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {A.3}相关工具}{417}{section.1.A.3}
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {A.3.1}数据预处理工具}{417}{subsection.1.A.3.1}
\defcounter {refsection}{0}\relax
\contentsline {subsection}{\numberline {A.3.2}评价工具}{418}{subsection.1.A.3.2}
\defcounter {refsection}{0}\relax
\contentsline {chapter}{\numberline {B}附录B}{419}{Appendix.2.B}
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {B.1}IBM模型3训练方法}{419}{section.2.B.1}
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {B.2}IBM模型4训练方法}{421}{section.2.B.2}
\defcounter {refsection}{0}\relax
\contentsline {section}{\numberline {B.3}IBM模型5训练方法}{423}{section.2.B.3}
\contentsfinish
\contentsfinish
Book/structure.tex
查看文件 @
623389ab
...
@@ -48,6 +48,7 @@
...
@@ -48,6 +48,7 @@
\geometry
{
\geometry
{
paper=b5paper,
% Paper size, change to letterpaper for US letter size
paper=b5paper,
% Paper size, change to letterpaper for US letter size
%papersize={185mm,260mm}, % specify paper size by (width,height)
top=2cm,
% Top margin
top=2cm,
% Top margin
bottom=1.5cm,
% Bottom margin
bottom=1.5cm,
% Bottom margin
left=1.8cm,
% Left margin
left=1.8cm,
% Left margin
...
...
编写
预览
Markdown
格式
0%
重试
或
添加新文件
添加附件
取消
您添加了
0
人
到此讨论。请谨慎行事。
请先完成此评论的编辑！
取消
请
注册
或者
登录
后发表评论
