Work

Active Learning between a Robot Learner and a Human Teacher

Year: 2012

Type: conference-paper

Abstract: A relative new trend in machine learning is interactive learning.Rather than passively absorbing knowledge, a learning agent can shape its learning experience through interaction with its teacher.As such, the learning becomes a bidirectional exchange.In a series of experiments, both in simulation and using a robotic platform, we examine how a learner that actively tries to influence the learning session can become more effective.Results show how interactive learning leads to more effective and faster learning, as social cues provided by the robot are picked up by human teachers.In addition, we observed gender differences, with female interactants being significantly more receptive to social cues displayed by the robot.

Source: Ghent University Academic Bibliography (Ghent University)

Authors: Joachim de Greeff, Frédéric Delaunay, Tony Belpaeme

Institution: University of Plymouth

Language: English

Topic: Social Robot Interaction and HRI

Subfield: Social Psychology

Field: Psychology

Domain: Social Sciences

SDG: Quality education

Open Access status: green

Funders: European Commission, Engineering and Physical Sciences Research Council

Awards: Linguistic and direct transmission of concepts in robot-human networks, Adaptive Strategies for Sustainable Long-Term Social Interaction