Keith E. Mathias

Evaluating evolutionary algorithms

Test functions are commonly used to evaluate the effectiveness of different search algorithms. However, the results of evaluation are as dependent on the test problems as they are on the algorithms that are the subject of comparison. Unfortunately, developing a test suite for evaluating competing search algorithms is difficult without clearly defined evaluation goals. In this paper we discuss some basic principles that can be used to develop test suites and we examine the role of test suites as they have been used to evaluate evolutionary search algorithms. Current test suites include functions that are easily solved by simple search methods such as greedy hill-climbers. Some test functions also have undesirable characteristics that are exaggerated as the dimensionality of the search space is increased. New methods are examined for constructing functions with different degrees of nonlinearity, where the interactions and the cost of evaluation scale with respect to the dimensionality of the search space.

Lamarckian Evolution, The Baldwin Effect and Function Optimization

We compare two forms of hybrid genetic search. The first uses Lamarckian evolution, while the second uses a related method where local search is employed to change the fitness of strings, but the acquired improvements do not change the genetic encoding of the individual. The latter search method exploits the Baldwin effect. By modeling a simple genetic algorithm we show that functions exist where simple genetic algorithms without learning as well as Lamarckian evolution converge to the same local optimum, while genetic search utilizing the Baldwin effect converges to the global optimum. We also show that a simple genetic algorithm exploiting the Baldwin effect can sometimes outperform forms of Lamarckian evolution that employ the same local search strategy.

Test driving three 1995 genetic algorithms: New test functions and geometric matching

Genetic algorithms have attracted a good deal of interest in the heuristic search community. Yet there are several different types of genetic algorithms with varying performance and search characteristics. In this article we look at three genetic algorithms: an elitist simple genetic algorithm, the CHC algorithm and Genitor. One problem in comparing algorithms is that most test problems in the genetic algorithm literature can be solved using simple local search methods. In this article, the three algorithms are compared using new test problems that are not readily solved using simple local search methods. We then compare a local search method to genetic algorithms for geometric matching and examine a hybrid algorithm that combines local and genetic search. The geometric matching problem matches a model (e.g., a line drawing) to a subset of lines contained in a field of line fragments. Local search is currently the best known method for solving general geometric matching problems.

Changing representations during search: A comparative study of delta coding

Delta coding is an iterative genetic search strategy that dynamically changes the representation of the search space in an attempt to exploit different problem representations. Delta coding sustains search by reinitializing the population at each iteration of search. This helps to avoid the asymptotic performance typically observed in genetic search as the population becomes more homogeneous. Here, the optimization ability of delta coding is empirically compared against CHC, ESGA, GENITOR, and random mutation hill-climbing (RMHC) on a suite of well-known test functions with and without Gray coding. Issues concerning the effects of Gray coding on these test functions are addressed.

Staged hybrid genetic search for seismic data imaging

Seismic data interpretation problems are typically solved using computationally intensive local search methods which often result in inferior solutions. Here, a traditional hybrid genetic algorithm is compared with different staged hybrid genetic algorithms on the geophysical imaging static corrections problem. The traditional hybrid genetic algorithm used here applied local search to every offspring produced by genetic search. The staged hybrid genetic algorithms were designed to temporally separate the local and genetic search components into distinct phases so as to minimize interference between the two search methods. The results show that some staged hybrid genetic algorithms produce higher quality solutions while using significantly less computational time for this problem.<<ETX>>

Sequence Scheduling With Genetic Algorithms

The application of genetic algorithms to sequence scheduling problems grew out of attempts to use this method to solve Traveling Salesman Problems. A genetic recombination operator for the Traveling Salesman Problem which preserves adjacency (or edges between cities) was developed; this operator proved to be superior to previous genetic operators for this problem [15]. Recently, a new enhancement to the edge recombination operator has been developed which further improves performance when compared to the original operator. Using this operator in the context of the GENITOR algorithm we obtain best known solutions for 30 and 105 city problems with considerable consistency. Our first test of this approach to scheduling was optimization of a printed circuit production line at Hewlett Packard[16). Success with this problem led us to apply similar methods to production scheduling on a sequencing problem posed by the Coors Brewing Co. This work has resulted in new findings regarding sequencing operators and their emphasis on adjacency, order, and position.

The Effects of Control Parameters and Restarts on Search Stagnation in Evolutionary Programming

Previous studies concluded that the best performance from an evolutionary programming (EP) algorithm was obtained by tuning the parameters for each problem. These studies used fitness at a pre-specified number of evaluations as the criterion for measuring performance. This study uses a complete factorial design for a large set of parameters on a wider array of functions and uses the mean trials to find the global optimum when practical. Our results suggest that the most critical EP control parameter is the perturbation method/rate of the strategy variables that control algorithm search potential. We found that the decline of search capacity limits the difficulty of functions that can be successfully solved with EP. Therefore, we propose a soft restart mechanism that significantly improves EP performance on more difficult problems.

Noisy function evaluation and the delta-coding algorithm

Genetic algorithms are becoming increasingly popular as a tool for optimization in signal processing environments due to their tolerance for noise. Several types of genetic algorithms are compared against a mutation driven stochastic hill-climbing algorithm on a standard set of benchmark functions which have had Gaussian noise added to them. The genetic algorithms used in these comparisons include an elitist simple genetic algorithm, the CHC adaptive search algorithm, and delta coding. Finally several hybrid genetic algorithms are described and compared on a very large and noisy seismic data imaging problem.