Active choice of teachers, learning strategies and goals for a socially guided intrinsic motivation learner

TL;DR

This paper introduces SGIM-ACTS, an active learning framework for robots that strategically chooses learning strategies and teachers to efficiently acquire motor skills and generalize outcomes.

Contribution

It presents a hierarchical, goal-driven active learning architecture that dynamically selects learning strategies and teachers based on learning progress, improving efficiency over single-strategy methods.

Findings

SGIM-ACTS learns more efficiently than single strategies.

The system effectively selects the best strategy for each outcome.

It successfully generalizes to new outcomes.

Abstract

We present an active learning architecture that allows a robot to actively learn which data collection strategy is most efficient for acquiring motor skills to achieve multiple outcomes, and generalise over its experience to achieve new outcomes. The robot explores its environment both via interactive learning and goal-babbling. It learns at the same time when, who and what to actively imitate from several available teachers, and learns when not to use social guidance but use active goal-oriented self-exploration. This is formalised in the framework of life-long strategic learning. The proposed architecture, called Socially Guided Intrinsic Motivation with Active Choice of Teacher and Strategy (SGIM-ACTS), relies on hierarchical active decisions of what and how to learn driven by empirical evaluation of learning progress for each learning strategy. We illustrate with an experiment where a simulated robot learns to control its arm for realising two kinds of different outcomes. It has to choose actively and hierarchically at each learning episode: 1) what to learn: which outcome is most interesting to select as a goal to focus on for goal-directed exploration; 2) how to learn: which data collection strategy to use among self-exploration, mimicry and emulation; 3) once he has decided when and what to imitate by choosing mimicry or emulation, then he has to choose who to imitate, from a set of different teachers. We show that SGIM-ACTS learns significantly more efficiently than using single learning strategies, and coherently selects the best strategy with respect to the chosen outcome, taking advantage of the available teachers (with different levels of skills).