A Concept Learning Tool Based On Calculating Version Space Cardinality

TL;DR

This paper introduces VeSC-CoL, a novel concept learning method that uses version space cardinality with BDD and SAT to handle highly imbalanced datasets without cross-validation.

Contribution

The paper presents a new approach for concept learning that replaces cross-validation with version space cardinality, utilizing BDD and SAT for efficient computation.

Findings

Accurately learns target concepts on imbalanced datasets.

Effectively replaces cross-validation with version space cardinality.

Demonstrates computational feasibility with sufficient resources.

Abstract

In this paper, we proposed VeSC-CoL (Version Space Cardinality based Concept Learning) to deal with concept learning on extremely imbalanced datasets, especially when cross-validation is not a viable option. VeSC-CoL uses version space cardinality as a measure for model quality to replace cross-validation. Instead of naive enumeration of the version space, Ordered Binary Decision Diagram and Boolean Satisfiability are used to compute the version space. Experiments show that VeSC-CoL can accurately learn the target concept when computational resource is allowed.