Probabilistic task modelling for meta-learning

TL;DR

This paper introduces a probabilistic generative model for meta-learning tasks, combining variational auto-encoding and latent Dirichlet allocation to explicitly represent task themes and improve task selection.

Contribution

It presents a novel probabilistic task modeling approach that captures task uncertainty and relatedness, enhancing meta-learning performance.

Findings

Model accurately captures task uncertainty.

Task relatedness improves meta-learning efficiency.

Empirical results validate the model's effectiveness.

Abstract

We propose probabilistic task modelling -- a generative probabilistic model for collections of tasks used in meta-learning. The proposed model combines variational auto-encoding and latent Dirichlet allocation to model each task as a mixture of Gaussian distribution in an embedding space. Such modelling provides an explicit representation of a task through its task-theme mixture. We present an efficient approximation inference technique based on variational inference method for empirical Bayes parameter estimation. We perform empirical evaluations to validate the task uncertainty and task distance produced by the proposed method through correlation diagrams of the prediction accuracy on testing tasks. We also carry out experiments of task selection in meta-learning to demonstrate how the task relatedness inferred from the proposed model help to facilitate meta-learning algorithms.