Performance Evaluation of Genetic Algorithms and Evolutionary Programming in Optimization and Machine Learning

Optimization of a Can Size Problem Using Real Encoded Chromosome in Genetic Algorithm

“…For complex problems in which high degree of accuracy is required as far as optimal solution is concerned, these binary GAs need high computational time and memory [14] . Also, larger number of computation stages and processing time during chromosome encoding and other operations is required [12] . The type of encoding has a larger impact on performance of GA as well.…”

“…GA explores the possibility space by setting the feasibility solutions concurrently in order to search optimal or near-optimal solutions [8] . A GA code comprises chromosome encoding, population initialization, crossover and mutation operations [11,12] . Each individual string consist of some numbers, which is called as chromosome, it may or may not be a binary bit string and it generally represents a possible solution to the problem for which it is being coded [13] .…”

Genetic Algorithms and Genetic Programming

“…We have already proposed several new EA-variants. The first prototype of this new class of evolutionary search which considers the concept of controllable selection pressure ([Aff01c], [Aff02]) for information exchange between independently evolving subpopulations has been introduced with the Segregative Genetic Algorithm (SEGA) [Aff01a], [Aff01b]. Even if the SEGA is already able to produce very high quality results in terms of global solution quality, selection pressure has to be set by the user which is a very time con-suming and difficult challenge.…”