Two Projection Pursuit Algorithms for Machine Learning under Non-Stationarity

TL;DR

This paper introduces two linear projection algorithms designed to handle non-stationarity in machine learning, with applications demonstrated in Brain Computer Interfacing, enhancing robustness and detection of non-stationary data.

Contribution

It proposes novel projection algorithms specifically tailored for non-stationary data, improving classification robustness in challenging scenarios.

Findings

The first algorithm effectively identifies maximally non-stationary directions.

The second algorithm improves two-way classification robustness.

Successful application demonstrated in Brain Computer Interfacing.

Abstract

This thesis derives, tests and applies two linear projection algorithms for machine learning under non-stationarity. The first finds a direction in a linear space upon which a data set is maximally non-stationary. The second aims to robustify two-way classification against non-stationarity. The algorithm is tested on a key application scenario, namely Brain Computer Interfacing.