LGMLSep 13, 2022

Unsupervised representation learning with recognition-parametrised probabilistic models

arXiv:2209.05661v27 citationsh-index: 54
AI Analysis

This provides a flexible framework for discovering latent structure in observational data, which is incremental but useful for AI and related domains.

The paper tackles unsupervised learning by introducing recognition-parametrised models (RPMs), a semi-parametric approach that learns latent structures without explicit generative models, achieving effective results on high-dimensional data like image classification and spatiotemporal datasets.

We introduce a new approach to probabilistic unsupervised learning based on the recognition-parametrised model (RPM): a normalised semi-parametric hypothesis class for joint distributions over observed and latent variables. Under the key assumption that observations are conditionally independent given latents, the RPM combines parametric prior and observation-conditioned latent distributions with non-parametric observation marginals. This approach leads to a flexible learnt recognition model capturing latent dependence between observations, without the need for an explicit, parametric generative model. The RPM admits exact maximum-likelihood learning for discrete latents, even for powerful neural-network-based recognition. We develop effective approximations applicable in the continuous-latent case. Experiments demonstrate the effectiveness of the RPM on high-dimensional data, learning image classification from weak indirect supervision; direct image-level latent Dirichlet allocation; and recognition-parametrised Gaussian process factor analysis (RP-GPFA) applied to multi-factorial spatiotemporal datasets. The RPM provides a powerful framework to discover meaningful latent structure underlying observational data, a function critical to both animal and artificial intelligence.

Code Implementations2 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes