Skip to content
A collection of Python code and machine learning exercises
Python Jupyter Notebook Shell
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
FeedForward
Paperspace
RBM
doc
notebooks
tests
CircleRandomWalk.py
GMM.py
Hopfield.py
IsingModel.py
KMeans.py
LICENSE
MetropolisHastings.py
RBM.py
README.md

README.md

MachineLearning

This repository contains code samples and supporting documentation for the series on machine learning and AI on my blog www.leftasexercise.com. Specifically, you will find the following ressources here.

  • An implementation of a restricted Boltzmann machine (RBM) in the RBM subdirectory, supporting different algorithms (ordinary contrastive divergence and persistent contrastive divergence (PCD)) using pure Python executing on a CPU and TensorFlow executing on a GPU (PCD only at the moment)
  • A Python implementation of an Ising model simulation
  • A Hopfield network in Python
  • the directory FeedForward that contains sample code related to classical feed forward networks like logistic regression, multinomial regression and layered networks
  • various notebooks that I use on my blog for the sake of illustration
  • and some additional short samples that I refer to in my blob
  • finally a directory with additional documentation in LaTex, including material on Boltzmann machines, Backpropagation, more on Ising models, a short introduction to Markov chains and Markov chain Monte Carlo methods, the EM algorithm for Gaussian mixed models and a short introduction into statistical physics and thermodynamics

Of course all these implementations and documents have been created for educational purposes and are not intended for production use - feel free to play with them and modify them if needed. When you use or cite this code or the documentation, please add a reference to this site or my blog www.leftasexercise.com.

You can’t perform that action at this time.