Skip to content

Mathematica implementations of machine learning algorithms used for prediction and personalization.

License

Notifications You must be signed in to change notification settings

antononcube/MathematicaForPrediction

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This open source project is for Mathematica implementations of machine learning algorithms that are used or can be used for prediction and personalization systems. (For prediction and facilitation of the behaviour of users, customers, clients, etc.) 

The algorithms implementations are given in Mathematica package files ("*.m"). Explanations or presentations about the algorithms are given in Mathematica notebook files ("*.nb").

The original set of algorithms is:
1. k-means and bisecting k-means;
2. associative rules finding;
3. decision trees and random forests;
4. non-negative matrix factorization;
5. prefix trees;
6. naive Bayesian classifiers generator;
7. a framework for linear vector space representations of document collections;
8. an item-item recommender framework based on sparse linear algebra.

In the future are going to be added algorithms for 
9. quantile regression,
10. self-organizing maps,
11. hierarchical clustering,
12. n-gram language models.

Also in this repository are going to be placed example or demonstrations notebooks for different applications of the algorithms listed above.

There is a blog associated with this project: http://mathematicaforprediction.wordpress.com .

I have implemented and extensively used these algorithms in search and prediction engines work in the last 5.5 years.

Anton Antonov
04.07.2013, Florida, USA