🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.ai
Article #9194

This WED 1/12/98 TACART: Schaal on Robot Learning

#9194
From: L. Andrew Campbell
Date: Tue, 20 Jan 1998 00:00
64 lines
3885 bytes
Los Angeles (LA) Association for Computing Machinery (ACM)
Technical Activity Committee on Artificial Intelligence (TACART)

             TACART

TOPIC:       Robot Learning: The Next Generation

SPEAKER:     Prof. Stefan Schaal, USC

DATE:        21 January 1997 (Wednesday night)

TIME:        The talk will begin at 7:30.

PLACE:       Loyola Marymount University, Pereira Hall.
             LMU is located at Loyola Blvd. and 80th Street in
             Westchester. From the 405, take the Manchester exit
             and proceed west to Loyola Blvd.  Turn north, and
             continue until you enter the LMU campus.  Ask the
             guard for directions to Pereira Hall. Parking is free.

ABSTRACT:    The next generation of successful robots will be very 
             different from those of today. Robots must learn
             continuously from their experiences, quickly react
             to a changing environment, and handle multiple
             tasks. In order to achieve these goals, a new
             approach to robotics is required: we need to program
             robots in the same way we "program" humans. When
             humans learn a task, they observe the task by
             somebody else, learn continuously through trial and
             error, and are instructed  about what is important
             and what is incorrect. Thus, robots have to learn,
             have to be taught, and have to be permitted to fail
             in a similar way.
             
             This talk will focus on three critical ingredients for 
             this new pproach: neural network learning algorithms for 
             continual incremental learning, nonlinear dynamic systems 
             for finding good planning and execution primitives, and 
             learning from demonstration. Videotape illustrations will
             show how results from our research have enabled a complex
             anthropomorphic robot arm to efficiently learn several
             manipulation tasks both autonomously and from a teacher.

BIOGRAPHY:   Stefan Schaal is an Assistant Professor at the Dept. 
             of Computer Science and the Neuroscience Program at
             USC. He holds additional appointments as Head of the
             Computational Learning Group of the Kawato Dynamic
             Brain Project (ERATO/JST) and at the Kinesiology 
             Dept. at Penn State. Before joining USC, Dr. Schaal
             was a postdoctoral fellow at MIT, held a position
             in Japan as Invited Researcher, and was an Adjunct
             Assistant Professor at the Georgia Tech.

DINNER:      No dinner is planned.  No social activities are planned.
NEWSGROUPS:  Announced in comp.ai, la.seminars, ca.seminars.
HOMEPAGE:    http://server.eng.lmu.edu/tacart/tacart.html
CONTACTS:    L. Andrew Campbell campbell@aero.org (310) 336-8642 or
             Program Chair Barney Toma tomab@ecn.com (310) 323-8451.
-- 
 zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
| L. Andrew Campbell     internet: campbell@aero.org       |
| M1-102 PO Box 92957    organization: The Aerospace Corp. |
| Los Angeles CA 90009   telephone: (310) 336-8642         |
Technical Details
Message-ID: <6a2oeo$mbq@news.aero.org>
Path: arc1-ca.pugleaf.net!unknown!not-for-mail

Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 10e37c,91e6fd3a71fdf634,start
X-Google-Attributes: gid10e37c,public
X-Google-Thread: 109296,91e6fd3a71fdf634,start
X-Google-Attributes: gid109296,public
X-Google-Thread: 10461e,91e6fd3a71fdf634,start
X-Google-Attributes: gid10461e,public
From: campbell@amanuensis.aero.org (L. Andrew Campbell)
Subject: This WED 1/12/98 TACART: Schaal on Robot Learning
Date: 1998/01/20
Message-ID: <6a2oeo$mbq@news.aero.org>
X-Deja-AN: 317789857
Distribution: ca
Organization: The Aerospace Corporation
Newsgroups: comp.ai,ca.seminars,la.seminars