Linear combination of one-step predictive information with an external reward in an episodic policy gradient setting: a critical analysis
Open Access
- 1 January 2013
- journal article
- research article
- Published by Frontiers Media SA in Frontiers in Psychology
- Vol. 4, 60129
Abstract
One of the main challenges in the field of embodied artificial intelligence is the open-ended autonomous learning of complex behaviors. Our approach is to use task-independent, information-driven intrinsic motivation(s) to support task-dependent learning. The work presented here is a preliminary step in which we investigate the predictive information (the mutual information of the past and future of the sensor stream) as an intrinsic drive, ideally supporting any kind of task acquisition. Previous experiments have shown that the predictive information (PI) is a good candidate to support autonomous, open-ended learning of complex behaviors, because a maximization of the PI corresponds to an exploration of morphology- and environment-dependent behavioral regularities. The idea is that these regularities can then be exploited in order to solve any given task. Three different experiments are presented and their results lead to the conclusion that the linear combination of the one-step PI with an external reward function is not generally recommended in an episodic policy gradient setting. Only for hard tasks a great speed-up can be achieved at the cost of an asymptotic performance lost.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Information Driven Self-Organization of Complex Robotic BehaviorsPLOS ONE, 2013
- Learning and exploration in action-perception loopsFrontiers in Neural Circuits, 2013
- Variants of guided self-organization for robot controlTheory in Biosciences, 2011
- Deriving neural network controllers from neuro-biological data: implementation of a single-leg stick insect controllerBiological Cybernetics, 2011
- Parameter-exploring policy gradientsNeural Networks, 2010
- Self-Organization, Embodiment, and Biologically Inspired RoboticsScience, 2007
- Reward, Motivation, and Reinforcement LearningNeuron, 2002
- Predictability, Complexity, and LearningNeural Computation, 2001
- Reinforcement Learning in Continuous Time and SpaceNeural Computation, 2000
- Inferring statistical complexityPhysical Review Letters, 1989