Efficient Representations for Life-Long Learning and Autoencoding

TL;DR

This paper introduces efficient algorithms for lifelong learning that develop internal representations to capture shared structures among tasks, enabling faster learning and better data efficiency.

Contribution

It proposes novel algorithms for learning shared low-dimensional structures and nonlinear Boolean combinations, advancing lifelong learning and autoencoding methods.

Findings

Algorithms effectively learn shared low-dimensional subspaces.

Methods successfully identify Boolean feature combinations.

Approach improves data efficiency in lifelong learning scenarios.

Abstract

It has been a long-standing goal in machine learning, as well as in AI more generally, to develop life-long learning systems that learn many different tasks over time, and reuse insights from tasks learned, "learning to learn" as they do so. In this work we pose and provide efficient algorithms for several natural theoretical formulations of this goal. Specifically, we consider the problem of learning many different target functions over time, that share certain commonalities that are initially unknown to the learning algorithm. Our aim is to learn new internal representations as the algorithm learns new target functions, that capture this commonality and allow subsequent learning tasks to be solved more efficiently and from less data. We develop efficient algorithms for two very different kinds of commonalities that target functions might share: one based on learning common low-dimensional and unions of low-dimensional subspaces and one based on learning nonlinear Boolean combinations of features. Our algorithms for learning Boolean feature combinations additionally have a dual interpretation, and can be viewed as giving an efficient procedure for constructing near-optimal sparse Boolean autoencoders under a natural "anchor-set" assumption.