Toy Machine Learning Algorithms

They were developed for didactic purposes. They do not implement all features of the original algorithms. They were made to be as much simple (and short) as possible. The only implementation goal was easy understandability.

The algorithms are implemented in YAP-prolog which is fairly standard so they should be easily adaptable to other prolog's.

• AQ algorithm (based on the AQ system from Michalski)
  • YAP-prolog Source Code
  • User's Guide
  • Example data set (in portuguese)
• ID3 algorithm (based on the ID3 system from Quinlan)
  • YAP-prolog Source Code
  • User's Guide
  • Example data set (in portuguese)
• RL algorithm (based on the YAILS system from Luis Torgo)
  • YAP-prolog Source Code
  • User's Guide
  • Example data set (in portuguese)