A Concept Learning Tool Based On Calculating Version Space Cardinality

2018/03/23 by Kuo-Kai Hsieh, Hsieh, Kuo-Kai, Li-C. Wang +1
Computer Science · #Machine Learning and Algorithms #Natural Language Processing Techniques #Software Engineering Research

paper · pdf · doi:10.48550/arxiv.1803.08625

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.

Citations

• Graph-Based Algorithms for Boolean Function Manipulation

Related

• Sample-efficient Learning of Concepts with Theoretical Guarantees: from Data to Concepts without Interventions
• Eliminating Information Leakage in Hard Concept Bottleneck Models with Supervised, Hierarchical Concept Learning
• Addressing Concept Mislabeling in Concept Bottleneck Models Through Preference Optimization
• V-CEM: Bridging Performance and Intervenability in Concept-based Models
• On the Concept Trustworthiness in Concept Bottleneck Models