This is a fork of stanford_dl_ex with the solutions for Stanford UFLDL Exercises.
Note that these are NOT the official solutions : they are simply solutions which I have coded and worked decently to give me reasonable solutions.
Note that this repository is made for Octave . Mose functionalities should work equally fine on Matlab, but I have not tested them on Matlab.
- Clone this repository
- Compile optimizers used throughout this code base:
cd path/to/repository/stanford_dl_ex/common/minFunc_2012
octave mexAll.m
# This will start some prerequiste files
- Install the io and statistics package (used for
randsample
in gradient checking subroutines)
# in Any Directory
octave # enter the octave Prompt
octave:1> pkg install -forge io
octave:1> pkg install -forge statistics
In addition to solving the problems, I have made some changes to adapt the codebase for Octave.
- Change
mex -outdir ...
intomex -o ...
in mexAll.m - Manually load statistics package in
grad_check.m