Saved in:
Bibliographic Details
Main Authors: Sikar, Daniel, Garcez, Artur, Bloomfield, Robin, Weyde, Tillman, Peeroo, Kaleem, Singh, Naman, Hutchinson, Maeve, Laksono, Dany, Reljan-Delaney, Mirela
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.07818
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911985748148224
author Sikar, Daniel
Garcez, Artur
Bloomfield, Robin
Weyde, Tillman
Peeroo, Kaleem
Singh, Naman
Hutchinson, Maeve
Laksono, Dany
Reljan-Delaney, Mirela
author_facet Sikar, Daniel
Garcez, Artur
Bloomfield, Robin
Weyde, Tillman
Peeroo, Kaleem
Singh, Naman
Hutchinson, Maeve
Laksono, Dany
Reljan-Delaney, Mirela
contents This study introduces the Misclassification Likelihood Matrix (MLM) as a novel tool for quantifying the reliability of neural network predictions under distribution shifts. The MLM is obtained by leveraging softmax outputs and clustering techniques to measure the distances between the predictions of a trained neural network and class centroids. By analyzing these distances, the MLM provides a comprehensive view of the model's misclassification tendencies, enabling decision-makers to identify the most common and critical sources of errors. The MLM allows for the prioritization of model improvements and the establishment of decision thresholds based on acceptable risk levels. The approach is evaluated on the MNIST dataset using a Convolutional Neural Network (CNN) and a perturbed version of the dataset to simulate distribution shifts. The results demonstrate the effectiveness of the MLM in assessing the reliability of predictions and highlight its potential in enhancing the interpretability and risk mitigation capabilities of neural networks. The implications of this work extend beyond image classification, with ongoing applications in autonomous systems, such as self-driving cars, to improve the safety and reliability of decision-making in complex, real-world environments.
format Preprint
id arxiv_https___arxiv_org_abs_2407_07818
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle The Misclassification Likelihood Matrix: Some Classes Are More Likely To Be Misclassified Than Others
Sikar, Daniel
Garcez, Artur
Bloomfield, Robin
Weyde, Tillman
Peeroo, Kaleem
Singh, Naman
Hutchinson, Maeve
Laksono, Dany
Reljan-Delaney, Mirela
Machine Learning
This study introduces the Misclassification Likelihood Matrix (MLM) as a novel tool for quantifying the reliability of neural network predictions under distribution shifts. The MLM is obtained by leveraging softmax outputs and clustering techniques to measure the distances between the predictions of a trained neural network and class centroids. By analyzing these distances, the MLM provides a comprehensive view of the model's misclassification tendencies, enabling decision-makers to identify the most common and critical sources of errors. The MLM allows for the prioritization of model improvements and the establishment of decision thresholds based on acceptable risk levels. The approach is evaluated on the MNIST dataset using a Convolutional Neural Network (CNN) and a perturbed version of the dataset to simulate distribution shifts. The results demonstrate the effectiveness of the MLM in assessing the reliability of predictions and highlight its potential in enhancing the interpretability and risk mitigation capabilities of neural networks. The implications of this work extend beyond image classification, with ongoing applications in autonomous systems, such as self-driving cars, to improve the safety and reliability of decision-making in complex, real-world environments.
title The Misclassification Likelihood Matrix: Some Classes Are More Likely To Be Misclassified Than Others
topic Machine Learning
url https://arxiv.org/abs/2407.07818