Saved in:
Bibliographic Details
Main Authors: Yang, Qidong, Zhu, Weicheng, Keslin, Joseph, Zanna, Laure, Rudner, Tim G. J., Fernandez-Granda, Carlos
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.23272
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913568096518144
author Yang, Qidong
Zhu, Weicheng
Keslin, Joseph
Zanna, Laure
Rudner, Tim G. J.
Fernandez-Granda, Carlos
author_facet Yang, Qidong
Zhu, Weicheng
Keslin, Joseph
Zanna, Laure
Rudner, Tim G. J.
Fernandez-Granda, Carlos
contents Probabilistic prediction of sequences from images and other high-dimensional data is a key challenge, particularly in risk-sensitive applications. In these settings, it is often desirable to quantify the uncertainty associated with the prediction (instead of just determining the most likely sequence, as in language modeling). In this paper, we propose a Monte Carlo framework to estimate probabilities and confidence intervals associated with the distribution of a discrete sequence. Our framework uses a Monte Carlo simulator, implemented as an autoregressively trained neural network, to sample sequences conditioned on an image input. We then use these samples to estimate the probabilities and confidence intervals. Experiments on synthetic and real data show that the framework produces accurate discriminative predictions, but can suffer from miscalibration. In order to address this shortcoming, we propose a time-dependent regularization method, which is shown to produce calibrated predictions.
format Preprint
id arxiv_https___arxiv_org_abs_2410_23272
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Monte Carlo Framework for Calibrated Uncertainty Estimation in Sequence Prediction
Yang, Qidong
Zhu, Weicheng
Keslin, Joseph
Zanna, Laure
Rudner, Tim G. J.
Fernandez-Granda, Carlos
Machine Learning
Artificial Intelligence
Probabilistic prediction of sequences from images and other high-dimensional data is a key challenge, particularly in risk-sensitive applications. In these settings, it is often desirable to quantify the uncertainty associated with the prediction (instead of just determining the most likely sequence, as in language modeling). In this paper, we propose a Monte Carlo framework to estimate probabilities and confidence intervals associated with the distribution of a discrete sequence. Our framework uses a Monte Carlo simulator, implemented as an autoregressively trained neural network, to sample sequences conditioned on an image input. We then use these samples to estimate the probabilities and confidence intervals. Experiments on synthetic and real data show that the framework produces accurate discriminative predictions, but can suffer from miscalibration. In order to address this shortcoming, we propose a time-dependent regularization method, which is shown to produce calibrated predictions.
title A Monte Carlo Framework for Calibrated Uncertainty Estimation in Sequence Prediction
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2410.23272