https://github.com/emaadmanzoor/python-ml-minimal

Linear and logistic regression in Python with no dependencies on Scipy/Numpy.
https://github.com/emaadmanzoor/python-ml-minimal

Last synced: about 1 month ago
JSON representation

Linear and logistic regression in Python with no dependencies on Scipy/Numpy.

• Host: GitHub
• URL: https://github.com/emaadmanzoor/python-ml-minimal
• Owner: emaadmanzoor
• Created: 2012-04-23T02:35:44.000Z (over 12 years ago)
• Default Branch: master
• Last Pushed: 2012-04-23T02:36:20.000Z (over 12 years ago)
• Last Synced: 2024-04-16T01:44:38.909Z (8 months ago)
• Language: Python
• Size: 86.9 KB
• Stars: 6
• Watchers: 3
• Forks: 5
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
Python ML Minimal

=================

There are some programming situations where one

might not be allowed to use Numpy/Scipy and is

required to perform some simple regression or

classification tasks; programming competitions

are one example.

The code here has no dependencies that are not

in the Python 2.x core, so they may be used in

these situations with ease. Do note that not

using Numpy/Scipy optimization libraries results

in a decrease in both efficiency and accuracy;

do not use this code unless your hands are tied.

## Contents

   1. linear-regression-one-var: Univariate

      linear regression with feature scaling.

   2. linear-regression-mult-var: Multivariate

      linear regression with feature scaling.

   3. logistic-regression: Logistic regression

      with feature scaling.

## Credits

   * All the algorithms here have been derived from

     Octave code I wrote for the [ML-Class](http://www.ml-class.org)

     programming assignments.

   * Very useful insights and workarounds for issues

     with machine learning in minimal Python have been

     lifted from [OT's blog](http://nestedinfiniteloops.wordpress.com/).