How it Works

This page gives a short and rather general overview over how tha LearningFramework is designed, how it works under the hood and how it can best be used.

More detailled notes about implementation details and technical details are in the Developer Wiki

Relevant terms and concepts

Note:

Examples:

Here is how some of the concepts/terms above relate to GATE concepts:

Learning algorithms included in the LearningFramework

The LearningFramework provides a range of algorithms, usually several different algorithms for a learning taks. Currently the following groups of algorithms are available:

Representation of instances

For training, the LearningFramework processes each document and creates an internal representation of each instance from the features extracted from the document. Which internal representation is used depends on the learning algorithm used.

Currently the following representations are implemented: