A new data structure HC-expression for learning from examples

摘要

A new data structure Hierarchical Counterfactual Expression (HC-Expression) is proposed. Its use in the area of learning from examples is studied. HC-Expression is a tree-like structure with alternate levels representing positive and negative exceptions to the rule. It is flexible and powerful enough to describe disjunctive concepts and can be visualised as a decision tree. Expressions to describe a concept can be efficiently generated from a set of examples and counter examples of a concept. The efficacy of the proposed method is examined by applying it to a set of data collected from the Institute of Indian Medicine.