TRESTLE: A Model of Concept Formation in Structured Domains

TL;DR

TRESTLE is an incremental probabilistic model that unifies various concept formation characteristics, creating hierarchical categories capable of handling mixed data types, and performs competitively with nonincremental models and humans.

Contribution

It introduces TRESTLE, a comprehensive hierarchical concept formation model that accounts for multiple human-like learning features in structured domains.

Findings

TRESTLE performs well on supervised and unsupervised tasks.

It closely matches human categorization behavior.

It outperforms nonincremental models in key metrics.

Abstract

The literature on concept formation has demonstrated that humans are capable of learning concepts incrementally, with a variety of attribute types, and in both supervised and unsupervised settings. Many models of concept formation focus on a subset of these characteristics, but none account for all of them. In this paper, we present TRESTLE, an incremental account of probabilistic concept formation in structured domains that unifies prior concept learning models. TRESTLE works by creating a hierarchical categorization tree that can be used to predict missing attribute values and cluster sets of examples into conceptually meaningful groups. It updates its knowledge by partially matching novel structures and sorting them into its categorization tree. Finally, the system supports mixed-data representations, including nominal, numeric, relational, and component attributes. We evaluate TRESTLE's performance on a supervised learning task and an unsupervised clustering task. For both tasks, we compare it to a nonincremental model and to human participants. We find that this new categorization model is competitive with the nonincremental approach and more closely approximates human behavior on both tasks. These results serve as an initial demonstration of TRESTLE's capabilities and show that, by taking key characteristics of human learning into account, it can better model behavior than approaches that ignore them.