On Perfect Classification and Clustering for Gaussian Processes

TL;DR

This paper introduces a data transformation for Gaussian processes that guarantees perfect separation in classification and clustering tasks, with proven asymptotic properties and demonstrated empirical success.

Contribution

It proposes a novel transformation for Gaussian processes that ensures perfect classification and clustering separation, supported by theoretical and empirical analysis.

Findings

Misclassification probability approaches zero asymptotically.

Transformation achieves perfect separation in clustering.

Method outperforms existing parametric and nonparametric techniques.

Abstract

In this paper, we propose a data based transformation for infinite-dimensional Gaussian processes and derive its limit theorem. For a classification problem, this transformation induces complete separation among the associated Gaussian processes. The misclassification probability of any simple classifier when applied on the transformed data asymptotically converges to zero. In a clustering problem using mixture models, an appropriate modification of this transformation asymptotically leads to perfect separation of the populations. Theoretical properties are studied for the usual $k$-means clustering method when used on this transformed data. Good empirical performance of the proposed methodology is demonstrated using simulated as well as benchmark data sets, when compared with some popular parametric and nonparametric methods for such functional data.