Decision Trees (Learning)
You are in the office pool, currently betting on the outcome of
the basketball game next week, between the MallRats and the Chinooks. You
have to decide which team will win, then bet on that team.
Of course, you could just guess,or flip a coin. Here we present a way that
(typically) will do better: by using observations about the past
performance of the teams.
Flash back to 1980. A small boy, no more than 12 years old, stands nervously, awaiting the start of the competition. A small multi-coloured cube is in his hands. The fingers twitch with anticipation, kneading the cube as if it were bread dough. The judges give their signal and… three… two… one… Go! All traces of nervousness are gone, as the boy shows amazing dexterity. The cube dances in his hands as he nimbly twists and rotates parts of the cube. Bang! The cube is thrust on to the table, stopping the clock. Twenty-seven seconds have elapsed since he started, but that is too long. To be a champion Rubik’s Cube solver, one has to be work faster — 20 seconds is competitive; 27 is not. With a look of disappointment, the boy leaves the stage, and another comes forward to take his place....
Overview of Project
For more information, contact
Dep't of Computing Science
University of Alberta
Edmonton, AB T6G 2H1 Canada
FAX: (780) 492-1071