Which is the best intrinsic motivation signal for learning multiple skills?
| Autor: | Vieri Giuliano Santucci, Gianluca eBaldassarre, Marco eMirolli |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
reinforcement learning
Computer science Biomedical Engineering 02 engineering and technology Kinematics lcsh:RC321-571 competence acquisition 03 medical and health sciences 0302 clinical medicine Artificial Intelligence multiple skills 0202 electrical engineering electronic engineering information engineering Intrinsic motivation Reinforcement learning Original Research Article Reinforcement lcsh:Neurosciences. Biological psychiatry. Neuropsychiatry Competence (human resources) business.industry hierarchical architecture Animal learning intrinsic motivations learning signals 020201 artificial intelligence & image processing Artificial intelligence Cache business Robotic arm 030217 neurology & neurosurgery simulated robot Neuroscience |
| Zdroj: | Frontiers in Neurorobotics Frontiers in Neurorobotics, Vol 7 (2013) Frontiers in neurorobotics 7 (2013): 22. doi:10.3389/fnbot.2013.00022 info:cnr-pdr/source/autori:Santucci, Vieri G; Baldassarre, Gianluca; Mirolli, Marco/titolo:Which is the best intrinsic motivation signal for learning multiple skills?/doi:10.3389%2Ffnbot.2013.00022/rivista:Frontiers in neurorobotics/anno:2013/pagina_da:22/pagina_a:/intervallo_pagine:22/volume:7 |
| ISSN: | 1662-5218 |
| DOI: | 10.3389/fnbot.2013.00022 |
| Popis: | Humans and other biological agents are able to autonomously learn and cache different skills in the absence of any biological pressure or any assigned task. In this respect, Intrinsic Motivations (i.e., motivations not connected to reward-related stimuli) play a cardinal role in animal learning, and can be considered as a fundamental tool for developing more autonomous and more adaptive artificial agents. In this work, we provide an exhaustive analysis of a scarcely investigated problem: which kind of IM reinforcement signal is the most suitable for driving the acquisition of multiple skills in the shortest time? To this purpose we implemented an artificial agent with a hierarchical architecture that allows to learn and cache different skills. We tested the system in a setup with continuous states and actions, in particular, with a kinematic robotic arm that has to learn different reaching tasks. We compare the results of different versions of the system driven by several different intrinsic motivation signals. The results show (a) that intrinsic reinforcements purely based on the knowledge of the system are not appropriate to guide the acquisition of multiple skills, and (b) that the stronger the link between the IM signal and the competence of the system, the better the performance. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|