abstract = "Simple implementation of genetic programming by making
use of the column tables is discussed. Implementations
of Koza's genetic programming in compiled languages are
usually not most efficient when crossover is applied.
If chromosomes are directed acyclic graphs, more
efficient than rooted trees both in memory requirement
as well as in evaluation time of chromosome, then
crossover requires traversing the data structures and
their preliminary analysis. Column tables inherently
code directed acyclic graphs, the implementation of
crossover is simple and needs neither traversing nor
checking of integrity of resulting data structures and
should be therefore more efficient. Stochastic
transformation operation mutation is also easily
defined. Column tables can represent graphs with
several output nodes and may be used e.g. for
optimization of feed-forward neural networks. Simple
illustrative examples of symbolic regression based on
the column tables are presented.",
notes = "WSC2 Second On-line World Conference on Soft Computing
in Engineering Design and
Manufacturing