| The goal of this book is to develop efficient optimization algorithms to solve diverse real-world problems of graded difficulty. Genetic and evolutionary mechanisms have been deployed for reaching the goal.
This book has made five significant contributions in the realm of genetic and evolutionary computation (GEC).
Practical guidelines for developing genetic algorithms (GAs) to solve realworld problems have been proposed. This fills a long standing gap between theory and practice of GAs. A practical population-sizing model for computing solutions with desired quality has also been developed. The model needs no statistical information about the problems. It has duly been validated by computer simulation experiments.
The suggested design-guidelines have been followed in developing a GA for solving the shortest path (SP) routing problem. Experimental studies validate the effectiveness of the guidelines. Further, the population-sizing model passes the feasibility test for this application. It appears to be applicable to a wide class of problems.
Elitist compact genetic algorithms (cGAs) have been developed under the framework of simple estimation of distribution algorithms (EDAs). They can deal with memory- and time-constrained problems. In addition, they do not require any prior knowledge about the problems. The design approach enables a typical cGA to overcome selection noise. This is achieved by persisting with the current best solution until, hopefully a better solution is found. A higher quality of solutions and a higher rate of convergence are attained in this way for most of the test problems. The hidden connection between EDAs and evolutionary strategies (ESs) has been made explicit. An analytical justification of this relationship is followed by its empirical verification. Further, a speedup model that quantifies convergence improvement has also been developed. Experimental evidence has been supplied to support the claims. |
|
|
 | | | | Mac OS X Leopard Killer TipsKiller Tips books are written with one goal in mind: to allow the reader to work faster and smarter. In other books, you’ll often find that the most useful information is found in sidebars, tips, and notes. In a Killer Tips book, there’s nothing to weed through: it’s all sidebars, tips, and notes! Here, Scott Kelby gives you... |
|
