Robotic Navigation with Human Brain Signals and Deep Reinforcement Learning

The use of robots as educational learning tools is quite extensive worldwide, yet it is rather limited in special education. In particular, the use of robots in the field of special education is under skepticism since robots are frequently believed to be expensive with limited capacity. The latter may change with the advent of social robots, which can be used in special education as affordable tools for delivering sophisticated stimuli to children with learning difficulties also due to preexisting conditions. Pilot studies occasionally demonstrate the effectiveness of social robots in specific domains. This chapter overviews the engagement of social robots in special education including the authors' preliminary work in this field; moreover, it discusses their proposal for potential future extensions involving more autonomous (i.e., intelligent) social robots as well as feedback from human brain signals.

Download Full-text

Use of higher order spectrum in characterizing nonlinear interactions in human brain signals

Proceedings of the International Conference on Advances in Computing, Communications and Informatics - ICACCI '12 ◽

10.1145/2345396.2345556 ◽

2012 ◽

Author(s):

Susant Kumar Jena ◽

Cuachy Pradhan ◽

N. M. Elango ◽

N. Pradhan

Keyword(s):

Human Brain ◽

Higher Order ◽

Nonlinear Interactions ◽

Brain Signals ◽

Order Spectrum

Download Full-text

Neural computations underlying inverse reinforcement learning in the human brain

eLife ◽

10.7554/elife.29718 ◽

2017 ◽

Vol 6 ◽

Cited By ~ 12

Author(s):

Sven Collette ◽

Wolfgang M Pauli ◽

Peter Bossaerts ◽

John O'Doherty

Keyword(s):

Reinforcement Learning ◽

Human Brain ◽

Model Comparison ◽

Formal Model ◽

Food Preferences ◽

Slot Machines ◽

Inverse Reinforcement Learning ◽

Neural Computations ◽

Fmri Analysis ◽

Imitation Strategy

In inverse reinforcement learning an observer infers the reward distribution available for actions in the environment solely through observing the actions implemented by another agent. To address whether this computational process is implemented in the human brain, participants underwent fMRI while learning about slot machines yielding hidden preferred and non-preferred food outcomes with varying probabilities, through observing the repeated slot choices of agents with similar and dissimilar food preferences. Using formal model comparison, we found that participants implemented inverse RL as opposed to a simple imitation strategy, in which the actions of the other agent are copied instead of inferring the underlying reward structure of the decision problem. Our computational fMRI analysis revealed that anterior dorsomedial prefrontal cortex encoded inferences about action-values within the value space of the agent as opposed to that of the observer, demonstrating that inverse RL is an abstract cognitive process divorceable from the values and concerns of the observer him/herself.

Download Full-text