Classification of Individuals with Complex Structure

Abstract

This paper introduces a foundation for inductive learning based on the use of higher-order logic for knowledge representation. In particular, the paper (i) provides a systematic individuals-as-terms approach to knowledge representation for inductive learning, and demonstrates the utility of types and higherorder constructs for this purpose; (ii) introduces a systematic way to construct predicates for use in induced definitions; and (iii) widens the applicability of decision-tree algorithms beyond the usual attribute-value setting to the classification of individuals with complex internal structure. The paper contains several illustrative applications. The effectiveness of the approach is demonstrated by applying the decision-tree learning system to two benchmark problems. 1. Introduction Traditionally, inductive learners have used the attribute-value language to represent individuals in (supervised) learning from examples. Though the relative simplicity of this attri...