This is another (work in progress) Chinese translation of Michael Nielsen's Neural Networks and Deep Learning, originally my learning notes of this free online book. It's written in LaTeX for better look and cross-referencing of math equations and plots. And I borrowed some finished work from https://github.com/tigerneil/neural-networks-and-deep-learning-zh-cn.
To compile the source code to a PDF file, please make sure you have a latest TeX system installed. You can download and install a TeX distribution for your platform from http://tug.org.
It's recommended to install
Also, please make sure you have these font packages installed: roboto
,
sourcecodepro
, sourceserifpro
, as these fonts are required in typesetting. You can install them
with tlmgr
command.
Note: If your TeX system is not installed in the default /usr/local
, you need to change the fontdir
macro in westernfonts.tex
, default is:
\newcommand{\fontdir}[0]{/usr/local/texlive/2015/texmf-dist/fonts/}
Use git to clone this repository and the code samples as a sub module:
$ git clone --recursive https://github.com/zhanggyb/nndl.git
or
$ git clone https://github.com/zhanggyb/nndl.git
$ cd nndl
$ git submodule update --init --recursive
Run the bootstrap.sh
in the working copy to download CJK fonts required:
$ ./bootstrap.sh
This is a simple shell script to download Google Noto CJK SC into fonts/
. You
can also manually download the
font file
and unarchive to fonts/
.
Run
$ make
to typeset and generate the PDF document. Or use your favorite GUI TeX application, but choose XeLaTeX as the typeset engine to support the chinese fonts.