Table of Contents

Pervasiveness of the problem of program induction; introduction to genetic algorithms; the representation problem for genetic algorithms; overview of genetic programming; detailed description of genetic programming; four introductory examples of genetic programming; amount of processing required to solve a problem; non-randomness of genetic programming; symbolic regression - error-driven evolution; control - cost-driven evolution; evolution of emergent behaviour; evolution of subsumption; entropy-driven evolution; evolution of strategy; co-evolution; evolution of classification; iteration, recursion, and setting; evolution of constrained syntactic structures; evolution of building blocks; evolution of hierarchies of building blocks; parallelization of genetic programming; ruggedness of genetic programming; extraneous variables and functions; operational issues; review of genetic programming; comparison with other paradigms; spontaneous emergence of self-replicating and evolutionarily self-improving computer programs. Appendices: computer implementation; problem-specific part of simple LISP code; kernel of the simple LISP code; embellishments to the simple LISP code; streamlined version of EVAL; editor for simplifying S-expressions; testing the simple LISP code; time-saving techniques; list of special symbols; list of special functions.

Promotional Information

John Koza has discovered a general and robust method of evolving computer programs that is effective over a breathtaking range of problems in applied mathematics, control engineering, and artificial intelligence. -- Stewart W. Wilson, The Rowland Institute for Science The research reported in this book is a tour de force. For the first time, since the idea was bandied about in the '40s and early '50s, we have a non-trivial, nontailored set of examples of automatic programming. -- John Holland, Professor of Psychology and Professor of Computer Science and Engineering, University of Michigan; External Professor, Santa Fe Institute

About the Author

John R. Koza is Consulting Associate Professor in the Computer Science Department at Stanford University.

Reviews

The research reported in this book is a tour de force. For the first time, since the idea was bandied about in the '40s and early '50s, we have a non-trivial, nontailored set of examples of automatic programming." John Holland