AI Expert
sign up for this
newsletter
2006
January
2005
December
•
November
•
October
•
September
•
August
•
July
•
June
•
May
•
April
•
March
•
February
•
January
2004
December
•
November
•
October
•
September
•
August
•
July
•
June
•
May
•
April
•
March
•
February
•
January
2003
December
•
November
•
October
•
September
August
•
July
•
June
•
May
April
•
March
•
February
• January
2002
December
• November
Issue #19
May 2004
By Dennis Merritt
AI - The art and science of making computers do interesting things
that are not in their nature.
Genetic Programming (GP)
The fun thing about AI is all the biological sorts of terms used to describe
cold programming algorithms, starting, of course, with the coining of
the phrase "artificial intelligence" to replace the rather boring
label "heuristic programming".
Genetic search algorithms follow this proud naming tradition. They are
also referred to has hill-climbing search algorithms. (See Jan 2003 newsletter
for discussion of genetic search and basketball schedules.) They explore
a search space by mutating a given point and seeing if any of the mutations
have higher values, in however height is measured, in the search space.
Genetic programming is the application of this same technique to the
generation of software.
Consider that there are all sorts of formal methodologies people use
for developing software. They are all good and have one thing in common.
They force a discipline on the programmer, requiring him/her to think
clearly and logically in the laying out of software to meet a given problem
specification.
There is another approach to developing software that is much more common
than one based on a formal methodology.
