Utilizing Explainable AI for improving the Performance of Neural Networks

TL;DR

This paper introduces a retraining pipeline that leverages explainable AI, specifically SHAP values, to assign weights to training data, resulting in improved neural network performance on real-world and public datasets.

Contribution

The paper presents a novel SHAP-based data weighting method for retraining neural networks, enhancing accuracy in tasks like people counting and image classification.

Findings

4% accuracy improvement on radar-based people counting

3% accuracy increase on CIFAR-10 dataset

Effective use of XAI for performance enhancement

Abstract

Nowadays, deep neural networks are widely used in a variety of fields that have a direct impact on society. Although those models typically show outstanding performance, they have been used for a long time as black boxes. To address this, Explainable Artificial Intelligence (XAI) has been developing as a field that aims to improve the transparency of the model and increase their trustworthiness. We propose a retraining pipeline that consistently improves the model predictions starting from XAI and utilizing state-of-the-art techniques. To do that, we use the XAI results, namely SHapley Additive exPlanations (SHAP) values, to give specific training weights to the data samples. This leads to an improved training of the model and, consequently, better performance. In order to benchmark our method, we evaluate it on both real-life and public datasets. First, we perform the method on a radar-based people counting scenario. Afterward, we test it on the CIFAR-10, a public Computer Vision dataset. Experiments using the SHAP-based retraining approach achieve a 4% more accuracy w.r.t. the standard equal weight retraining for people counting tasks. Moreover, on the CIFAR-10, our SHAP-based weighting strategy ends up with a 3% accuracy rate than the training procedure with equal weighted samples.