Uncertainty Prediction for Machine Learning Models of Material Properties

TL;DR

This paper compares three methods for estimating uncertainty in machine learning predictions of material properties, highlighting the advantages of directly modeling individual uncertainties for improved accuracy.

Contribution

It evaluates and compares three approaches to prediction interval estimation in ML models for material properties, providing insights into their relative strengths and weaknesses.

Findings

Direct modeling of uncertainties often yields better calibration.

Gaussian Processes offer a probabilistic framework for uncertainty estimation.

The study uses the JARVIS-DFT database for comprehensive testing.

Abstract

Uncertainty quantification in Artificial Intelligence (AI)-based predictions of material properties is of immense importance for the success and reliability of AI applications in material science. While confidence intervals are commonly reported for machine learning (ML) models, prediction intervals, i.e., the evaluation of the uncertainty on each prediction, are seldomly available. In this work we compare 3 different approaches to obtain such individual uncertainty, testing them on 12 ML-physical properties. Specifically, we investigated using the Quantile loss function, machine learning the prediction intervals directly and using Gaussian Processes. We identify each approachs advantages and disadvantages and end up slightly favoring the modeling of the individual uncertainties directly, as it is the easiest to fit and, in most cases, minimizes over-and under-estimation of the predicted errors. All data for training and testing were taken from the publicly available JARVIS-DFT database, and the codes developed for computing the prediction intervals are available through JARVIS-Tools.