Comments on "Challenges of cellwise outliers" by Jakob Raymaekers and Peter J. Rousseeuw

TL;DR

This paper discusses the challenges of cellwise outliers in robust statistics, reviewing recent methods and extending the discussion to time series data, emphasizing the importance of addressing this new outlier type.

Contribution

It provides a critical commentary on the recent work by Raymaekers and Rousseeuw, highlighting the significance of cellwise outliers and their treatment in various models, including time series.

Findings

Cellwise outliers are a significant challenge in robust statistics.

Recent methods have been developed to address cellwise contamination.

Time series data can also be affected by cellwise outliers.

Abstract

The main aim of robust statistics is the development of methods able to cope with the presence of outliers. A new type of outliers, namely "cellwise", has garnered considerable attention. The state of the art for dealing with cellwise contamination in different models is presented in Raymaekers and Rousseeuw (2024). Outliers in time series can be treated as cellwise outliers, a further discussion on this subject is presented.