A Self-Supervised Framework for Function Learning and Extrapolation

Abstract

Understanding how agents learn to generalize -- and, in particular, to extrapolate -- in high-dimensional, naturalistic environments remains a challenge for both machine learning and the study of biological agents. One approach to this has been the use of function learning paradigms, which allow peoples' empirical patterns of generalization for smooth scalar functions to be described precisely. However, to date, such work has not succeeded in identifying mechanisms that acquire the kinds of general purpose representations over which function learning can operate to exhibit the patterns of generalization observed in human empirical studies. Here, we present a framework for how a learner may acquire such representations, that then support generalization -- and extrapolation in particular -- in a few-shot fashion. Taking inspiration from a classic theory of visual processing, we construct a self-supervised encoder that implements the basic inductive bias of invariance under topological distortions. We show the resulting representations outperform those from other models for unsupervised time series learning in several downstream function learning tasks, including extrapolation.

arXiv comment15 pages, 4 figures

References in corpus (8)

• Distributed Representations of Words and Phrases and their Compositionality
• Auto-Encoding Variational Bayes
• Structure Discovery in Nonparametric Regression through Compositional Kernel Search
• Gaussian Process Kernels for Pattern Discovery and Extrapolation
• TimeNet: Pre-trained deep recurrent neural network for time series classification
• Unsupervised Representation Learning for Time Series with Temporal Neighborhood Coding
• Improving Clinical Predictions through Unsupervised Time Series Representation Learning
• Learning Representations that Support Extrapolation