Unsupervised Heterogeneous Coupling Learning for Categorical Representation

TL;DR

This paper introduces UNTIE, a novel unsupervised method for representing complex hierarchical and heterogeneous categorical data by disentangling couplings, leading to improved data representations without deep learning.

Contribution

The paper proposes UNTIE, a shallow yet effective unsupervised approach that captures heterogeneous couplings in categorical data, addressing limitations of existing methods and enhancing representation quality.

Findings

UNTIE outperforms state-of-the-art methods on 25 datasets.

It effectively disentangles heterogeneous couplings.

Theoretical analysis confirms maximal data separability.

Abstract

Complex categorical data is often hierarchically coupled with heterogeneous relationships between attributes and attribute values and the couplings between objects. Such value-to-object couplings are heterogeneous with complementary and inconsistent interactions and distributions. Limited research exists on unlabeled categorical data representations, ignores the heterogeneous and hierarchical couplings, underestimates data characteristics and complexities, and overuses redundant information, etc. The deep representation learning of unlabeled categorical data is challenging, overseeing such value-to-object couplings, complementarity and inconsistency, and requiring large data, disentanglement, and high computational power. This work introduces a shallow but powerful UNsupervised heTerogeneous couplIng lEarning (UNTIE) approach for representing coupled categorical data by untying the interactions between couplings and revealing heterogeneous distributions embedded in each type of couplings. UNTIE is efficiently optimized w.r.t. a kernel k-means objective function for unsupervised representation learning of heterogeneous and hierarchical value-to-object couplings. Theoretical analysis shows that UNTIE can represent categorical data with maximal separability while effectively represent heterogeneous couplings and disclose their roles in categorical data. The UNTIE-learned representations make significant performance improvement against the state-of-the-art categorical representations and deep representation models on 25 categorical data sets with diversified characteristics.