A comparative study of feature selection methods for stress hotspot classification in materials

TL;DR

This paper evaluates various feature selection methods for stress hotspot classification in materials, aiming to improve model accuracy and interpretability by identifying key microstructural features influencing stress buildup.

Contribution

It compares feature selection techniques in the context of stress hotspot classification, highlighting biases and recommending preferred methods for physical interpretation.

Findings

Certain feature selection methods are biased in this context.

A preferred technique for feature ranking enhances physical interpretability.

Insights into microstructural features influencing stress hotspots.

Abstract

The first step in constructing a machine learning model is defining the features of the data set that can be used for optimal learning. In this work we discuss feature selection methods, which can be used to build better models, as well as achieve model interpretability. We applied these methods in the context of stress hotspot classification problem, to determine what microstructural characteristics can cause stress to build up in certain grains during uniaxial tensile deformation. The results show how some feature selection techniques are biased and demonstrate a preferred technique to get feature rankings for physical interpretations.