合并分支 'caorunzhe' 到 'master'

Caorunzhe

查看合并请求 !393

Chapter16/chapter16.tex

@@ -42,7 +42,7 @@
 %----------------------------------------------------------------------------------------
 \subsection{数据增强}
 
-\parinterval {\small\bfnew{数据增强}}（Data Augmentation）是一种增加训练数据的方法，通常通过对既有数据进行修改或者生成新的伪数据等方式实现。有时候，数据增强也可以被看做是一种防止模型过拟合的手段（{\color{red} ref: A survey on Image Data Augmentation for Deep Learning}）。在机器翻译中，典型的数据增强方法包括回译、加噪和双语数据挖掘等。
+\parinterval {\small\bfnew{数据增强}}（Data Augmentation）是一种增加训练数据的方法，通常通过对既有数据进行修改或者生成新的伪数据等方式实现。有时候，数据增强也可以被看做是一种防止模型过拟合的手段\cite{DBLP:journals/jbd/ShortenK19}。在机器翻译中，典型的数据增强方法包括回译、加噪和双语数据挖掘等。
 
 \subsubsection{1. 回译}
 

bibliography.bib

@@ -6507,6 +6507,15 @@ author    = {Yoshua Bengio and
   publisher = {Annual Meeting of the Association for Computational Linguistics},
   year      = {2018}
 }
+@article{DBLP:journals/jbd/ShortenK19,
+  author    = {Connor Shorten and
+               Taghi M. Khoshgoftaar},
+  title     = {A survey on Image Data Augmentation for Deep Learning},
+  journal   = {J. Big Data},
+  volume    = {6},
+  pages     = {60},
+  year      = {2019}
+}
 %%%%% chapter 16------------------------------------------------------
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%