Functional Classwise Principal Component Analysis: A Novel Classification Framework

TL;DR

This paper introduces a new classification framework using classwise functional PCA for high-dimensional time series data, effectively addressing small sample size issues and improving classification in diverse fields.

Contribution

The paper proposes a novel functional data classification method utilizing classwise PCA, suitable for high-dimensional, small sample size time series data, and demonstrates its effectiveness across various domains.

Findings

Effective in high-dimensional, small sample size scenarios

Applicable to diverse fields like neuroscience and medical sciences

Improves classification accuracy with functional PCA

Abstract

In recent times, functional data analysis (FDA) has been successfully applied in the field of high dimensional data classification. In this paper, we present a novel classification framework using functional data and classwise Principal Component Analysis (PCA). Our proposed method can be used in high dimensional time series data which typically suffers from small sample size problem. Our method extracts a piece wise linear functional feature space and is particularly suitable for hard classification problems.The proposed framework converts time series data into functional data and uses classwise functional PCA for feature extraction followed by classification using a Bayesian linear classifier. We demonstrate the efficacy of our proposed method by applying it to both synthetic data sets and real time series data from diverse fields including but not limited to neuroscience, food science, medical sciences and chemometrics.