AI Forum: Andrew Barto/University of Massachussetts at Amherst Intrinsic Motivation and Computational Reinforcement Learning in ACES 2.402

Contact Name: 
Jenna Whitney
Feb 17, 2006 3:00pm - 4:00pm

There is a signup schedule for this event.


er Name/Affiliation: Andrew Barto/University of Massachussetts at Amherst<

Talk Title: Intrinsic Motivation and Computational Reinforcement Le


Date/Time: February 17 2006 at 3:00 p.m.

Coffee: 2:4

5 p.m.

Location: ACES 2.402

Host: Peter Stone

Talk A

Motivation is a key factor in human learning. We learn best
when we are highly motivated to learn. Psychologists distinguish

en extrinsically-motivated behavior which is behavior
undertaken to ac

hieve some externally supplied reward such
as a prize a high grade o

r a high-paying job and intrinsically-motivated
behavior which is beh

avior done for its own sake. Is there
an analogous distinction for mach

ine learning systems? Can
we say of a machine learning system that it i

s motivated
to learn and if so can it be meaningful to distinguish between extrinsic and intrinsic motivation? Further is
intrinsic mot

ivation something that we as machine learning
researchers should care a

bout? In this talk I argue that
the answer to each to each of these qu

estions is yes.
After presenting a brief overview of the history of idea

related to intrinsic motivation in machine learning I describe

ome of our recent computational experiments that explore
these ideas wi

thin the framework of computational reinforcement
learning (RL). It is

a common perception that computational
RL only deals with extrinsic rew

ard because an RL agent
is typically seen as receiving reward signals o

nly from
its external environment. To the contrary however I argue that the computational RL framework is particularly well
suited for i

ncorporating principles of intrinsic motivation
and I present our view
that extending learning in this direction
is important for creating co

mpetent adaptive agents.

Speaker Bio:
Andrew Barto is Professor

of Computer Science University
of Massachusetts Amherst. He received

his B.S. with distinction
in mathematics from the University of Michiga

n in 1970
and his Ph.D. in Computer Science in 1975 also from the
University of Michigan. He joined the Computer Science Department
of t

he University of Massachusetts Amherst in 1977 as a
Postdoctoral Resear

ch Associate became an Associate Professor
in 1982 and has been a Ful

l Professor since 1991. He is
Co-Director of the Autonomous Learning La

boratory and a
core faculty member of the Neuroscience and Behavior Pro

of the University of Massachusetts. His research centers
on le

arning in natural and artificial systems and he has
studied machine le

arning algorithms since 1977 contributing
to the development of the co

mputational theory and practice
of reinforcement learning. His current

research centers
on models of motor learning and reinforcement learning
for real-time planning and control with specific interest
in autonomous mental development through intrinsically motivated behavior.