Python implementation of the Machine Learning exercises of an open course by Standford University by Andrew Ng available at coursera.org. https://www.coursera.org/learn/machine-learning
In order to get most out of this Python exercises implementation, it is recommended to first take course video lectures each lecture comes with an exercise description PDF and Octave skeleton code please go through that. Specially go through the Octave scripts such as ex1.m, ex2.m...,ex8_cofi.m, ex8.m etc, starting with exX_ where X is a digit. After each class session there... you can run and study related Python code implementation included in this project.
- Tested on Python 64 bit 3.5
- numpy for Python 64 bit 3.5
- scipy Python 64 bit 3.5
- It is suggested to install "Anaconda 64 bit Python 3.5" to get Python environment.
- $> python main.py
- Course test data files already included under "./data" directory.
- Usually each algorithm is in its own Python class.
- Supervised learning algorithms are under "./supervised" package.
- Unsupervised learning algorithm are under "./unsupervised" package.
- Neural Networks algorithms are under "./supervised/neural_networks" package.
- Some utility routines are under "./utils" package.
- Scripts that execute and test the algorithms are under "./exercises/coursera" package.
- main.py scripts starts the execution.
- "Anaconda 64 bit Python 3.5" was installed to get Python environment.
- Testing was done on Windows 7 environment but since it is Python other platforms should not be a problem.
- Development was done in Visual Studio 2013 hence a workspace "MachineLearning.sln" file is included but other platform users can ignore "MachineLearning.sln" and "MachineLearning.pyproj" files.