Robust principal components for irregularly spaced longitudinal data

TL;DR

This paper introduces robust principal component methods for irregularly spaced longitudinal data, effectively handling missing values and contamination, with two approaches: an MM-estimator-based method and a fast smoothing-based estimator.

Contribution

It develops novel robust PCA techniques tailored for irregularly spaced and contaminated longitudinal data, addressing missingness and outliers simultaneously.

Findings

The simple estimator outperforms competitors on complete data.

The MM estimator is effective for incomplete data.

Both methods demonstrate robustness in real and simulated experiments.

Abstract

Consider longitudinal data $x_{ij},$ with $i=1,...,n$ and $j=1,...,p_{i},$ where $x_{ij}$ is the $j-$th observation of the random function $X_{i}\left( .\right) $ observed at time $t_{j}.$ The goal of this paper is to develop a parsimonious representation of the data by a linear combination of a set of $q$ smooth functions $H_{k}\left( .\right) $ ($k=1,..,q)$ in the sense that $x_{ij}\approx\mu_{j}+\sum_{k=1}^{q}\beta_{ki}H_{k}\left( t_{j}\right) ,$ such that it fulfills three goals: it is resistant to atypical $X_{i}$'s ('case contamination'), it is resistant to isolated gross errors at some $t_{ij}$ ('cell contamination'), and it can be applied when some of the $x_{ij}$ are missing ('irregularly spaced' ---or 'incomplete'-- data). Two approaches will be proposed for this problem. One deals with the three goals stated above, and is based on ideas similar to MM-estimation (Yohai 1987). The other is a simple and fast estimator which can be applied to complete data with case- and cellwise contamination, and is based on applying a standard robust principal components estimate and smoothing the principal directions. Experiments with real and simulated data suggest that with complete data the simple estimator outperforms its competitors, while the MM estimator is competitive for incomplete data. Keywords: Principal components, MM-estimator, longitudinal .data, B-splines, incomplete data.