Speaker
E. Vaandering
(VANDERBILT UNIVERSITY)
Description
Genetic programming is a machine learning technique, popularized by Koza in
1992, in which computer programs which solve user-posed problems are
automatically discovered. Populations of programs are evaluated for their
fitness of solving a particular problem. New populations of ever increasing
fitness are generated by mimicking the biological processes underlying
evolution. These processes are principally genetic recombination, mutation,
and survival of the fittest.
Genetic programming has potential advantages over other machine learning
techniques such as neural networks and genetic algorithms in that the form of
the solution is not specified in advance and the program can grow as large as
necessary to adequately solve the posed problem.
This talk will give an overview and demonstration of the genetic programming
technique and show a successful application in high energy physics: the
automatic construction of an event filter for FOCUS which is more powerful than
the experiment's usual methods of event selection. We have applied this method
to the study of doubly Cabibbo suppressed decays of charmed hadrons ($D^+$,
$D_s^+$, and $\Lambda_c^+$).
Primary author
E. Vaandering
(VANDERBILT UNIVERSITY)
