1994 MSc in Artifical Intellegence
University of British Columbia
"Reinforcement Learning with soccer playing
robots"
Reinforcement learning in a large, partialy observable, multiagent domain
poses some unique problems. In the soccer testbed
the environment of the robots is constantly changing, and the goals and
actions of a robot need to be evaluated at a rate proportional with the
rate of change of the environment.; Further the environment of the
robot is such that the coresponding state space for the MDP is too large,
for a optimal policy to be efficiently computed.
Located in southern Alberta, the
City of Lethbridge
has one of Canada's leading Undergraduate universities.
It is well known for its architectural design (Named as the building most
like a bomb shelter) it was designed by Arthur Erikson.
one of the fun courses I took was an Archaeological dig at
Tel Miqne. of course
I also had to take all the prerequisite course first. The only thing I ever found, in
addition to tones of burnt pottery was an egyptian scarab.
American School of oriental research