C-SALT: Mining Class-Specific ALTerations in Boolean Matrix Factorization

TL;DR

This paper introduces C-SALT, a parameter-free Boolean matrix factorization method that identifies class-specific alterations, extending beyond common or class-specific factorizations, applicable to complex class-dependent data structures.

Contribution

The paper presents a novel class-specific Boolean matrix factorization approach that captures complex class dependencies and is parameter-free, expanding the scope of existing methods.

Findings

Effective in filtering class-specific structures in synthetic data

Validates model assumptions on real-world datasets

Applicable to datasets with complex class dependencies

Abstract

Given labeled data represented by a binary matrix, we consider the task to derive a Boolean matrix factorization which identifies commonalities and specifications among the classes. While existing works focus on rank-one factorizations which are either specific or common to the classes, we derive class-specific alterations from common factorizations as well. Therewith, we broaden the applicability of our new method to datasets whose class-dependencies have a more complex structure. On the basis of synthetic and real-world datasets, we show on the one hand that our method is able to filter structure which corresponds to our model assumption, and on the other hand that our model assumption is justified in real-world application. Our method is parameter-free.