Estimating Regression Predictive Distributions with Sample Networks
DOI:
https://doi.org/10.1609/aaai.v37i6.25948Keywords:
ML: Probabilistic Methods, CV: 3D Computer Vision, ML: Bayesian LearningAbstract
Estimating the uncertainty in deep neural network predictions is crucial for many real-world applications. A common approach to model uncertainty is to choose a parametric distribution and fit the data to it using maximum likelihood estimation. The chosen parametric form can be a poor fit to the data-generating distribution, resulting in unreliable uncertainty estimates. In this work, we propose SampleNet, a flexible and scalable architecture for modeling uncertainty that avoids specifying a parametric form on the output distribution. SampleNets do so by defining an empirical distribution using samples that are learned with the Energy Score and regularized with the Sinkhorn Divergence. SampleNets are shown to be able to well-fit a wide range of distributions and to outperform baselines on large-scale real-world regression tasks.Downloads
Published
2023-06-26
How to Cite
Harakeh, A., Hu, J. S. K., Guan, N., Waslander, S., & Paull, L. (2023). Estimating Regression Predictive Distributions with Sample Networks. Proceedings of the AAAI Conference on Artificial Intelligence, 37(6), 7830-7838. https://doi.org/10.1609/aaai.v37i6.25948
Issue
Section
AAAI Technical Track on Machine Learning I