We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
A moment-matching metric for latent variable generative models.
- Authors
Beaulac, Cédric
- Abstract
It is difficult to assess the quality of a fitted model in unsupervised learning problems. Latent variable models, such as variational autoencoders and Gaussian mixture models, are often trained with likelihood-based approaches. In the scope of Goodhart's law, when a metric becomes a target it ceases to be a good metric and therefore we should not use likelihood to assess the quality of the fit of these models. The solution we propose is a new metric for model comparison or regularization that relies on moments. The key idea is to study the difference between the data moments and the model moments using a matrix norm, such as the Frobenius norm. We first show how to use this new metric for model comparison and then for regularization. We show that our proposed metric is faster to compute and has a smaller variance than the commonly used procedure of drawing samples from the fitted distribution. We conclude this article with a demonstration of both applications and we discuss our findings and future work.
- Subjects
GAUSSIAN mixture models; LATENT variables; MATRIX norms; LEARNING problems
- Publication
Machine Learning, 2023, Vol 112, Issue 10, p3749
- ISSN
0885-6125
- Publication type
Article
- DOI
10.1007/s10994-023-06340-x