IBM®
Skip to main content
    Country/region [change]    Terms of use
 
 
 
    Home    Products    Services & solutions    Support & downloads    My account    

IBM Technical Journals
Special report: Celebrating 50 years of the IBM Journals
All topics > Fundamental Science and Mathematics >

Some studies in machine learning using the game of checkers

 Award plaque by A. L. Samuel

Two machine-learning procedures have been investigated in some detail using the game of checkers. Enough work has been done to verify the fact that a computer can be programmed so that it will learn to play a better game of checkers than can be played by the person who wrote the program. Furthermore, it can learn to do this in a remarkably short period of time (8 or 10 hours of machine-playing time) when given only the rules of the game, a sense of direction, and a redundant and incomplete list of parameters which are thought to have something to do with the game, but whose correct signs and relative weights are unknown and unspecified. The principles of machine learning verified by these experiments are, of course, applicable to many other situations.

Originally published:

IBM Journal of Research and Development, Volume 3, Issue 3, pp. 210-229 (1959).

Significance:

This highly cited paper described early principles of how machines could learn to play games and applied them to playing checkers on an IBM 704 computer. This was the first program of its kind and was a milestone in artificial intelligence programming.

The machine-learning principles developed in this work were steps toward the ultimate match between chess champion Gary Kasparov and the IBM Deep Blue computer which showed that machines could learn to play complex games and were capable of defeating even the most formidable human opponents. Today, machine learning has countless applications in fields that range from search engines and medical diagnosis to bioinformatics, stock market analysis, and robot locomotion.

Comments:


    About IBM    Privacy    Contact