Behavior Of Genetic Algorithm Research Articles

Recommendations on the Implementation of Genetic Algorithms for the Directed Evolution of Enzymes for Industrial Purposes.

In directed evolution (DE) the assessment of candidate enzymes and their modification is essential. In this study we have investigated genetic algorithms (GAs) in this context and conducted a systematic study of the behavior of GAs on 20 fitness landscapes (FLs) of varying complexity. This has allowed the tuning of the GAs to be explored. On the basis of this study, recommendations for the best GA settings to use for a GA-directed high-throughput experimental program (in which populations and the number of generations is necessarily low) are reported. The FLs were based upon simple linear models and were characterized by the behavior of the GA on the landscape as demonstrated by stall plots and the footprints and adhesion of candidate solutions, which highlighted local optima (LOs). In order to maximize progress of the GA and to reduce the chances of becoming stuck in a LO it was best to use: 1) a large number of generations, 2) high populations, 3) removal of duplicate sequences (clones), 4) double mutation, and 5) high selection pressure (the two best individuals go to the next generation), and 6) to consider using a designed sequence as the starting point of the GA run. We believe that these recommendations might be appropriate starting points for studies employing GAs within DE experiments.

A mathematical analysis of the long-run behavior of genetic algorithms for social modeling

We present a mathematical analysis of the long-run behavior of genetic algorithms (GAs) that are used for modeling social phenomena. Our analysis relies on commonly used mathematical techniques in the field of evolutionary game theory. We make a number of assumptions in our analysis, the most important one being that the mutation rate is positive but infinitely small. Given our assumptions, we derive results that can be used to calculate the exact long-run behavior of a GA. Using these results, the need to rely on computer simulations can be avoided. We also show that if the mutation rate is infinitely small the crossover rate has no effect on the long-run behavior of a GA. To demonstrate the usefulness of our mathematical analysis, we replicate a well-known study by Axelrod in which a GA is used to model the evolution of strategies in iterated prisoner’s dilemmas. The theoretically predicted long-run behavior of the GA turns out to be in perfect agreement with the long-run behavior observed in computer simulations. Also, in line with our theoretically informed expectations, computer simulations indicate that the crossover rate has virtually no long-run effect. Some general new insights into the behavior of GAs in the prisoner’s dilemma context are provided as well.

Static and Dynamic Selection Thresholds Governing the Accumulation of Information in Genetic Algorithms Using Ranked Populations

Mutation applied indiscriminately across a population has, on average, a detrimental effect on the accumulation of solution alleles within the population and is usually beneficial only when targeted at individuals with few solution alleles. Many common selection techniques can delete individuals with more solution alleles than are easily recovered by mutation. The paper identifies static and dynamic selection thresholds governing accumulation of information in a genetic algorithm (GA). When individuals are ranked by fitness, there exists a dynamic threshold defined by the solution density of surviving individuals and a lower static threshold defined by the solution density of the information source used for mutation. Replacing individuals ranked below the static threshold with randomly generated individuals avoids the need for mutation while maintaining diversity in the population with a consequent improvement in population fitness. By replacing individuals ranked between the thresholds with randomly selected individuals from above the dynamic threshold, population fitness improves dramatically. We model the dynamic behavior of GAs using these thresholds and demonstrate their effectiveness by simulation and benchmark problems.

Analysis of new niching genetic algorithms for finding multiple solutions in the job shop scheduling

In this paper the performance of the most recent multi-modal genetic algorithms (MMGAs) on the Job Shop Scheduling Problem (JSSP) is compared in term of efficacy, multi-solution based efficacy (the algorithm's capability to find multiple optima), and diversity in the final set of solutions. The capability of Genetic Algorithms (GAs) to work on a set of solutions allows us to reach different optima in only one run. Nevertheless, simple GAs are not able to maintain different solutions in the last iteration, therefore reaching only one local or global optimum. Research based on the preservation of the diversity through MMGAs has provided very promising results. These techniques, known as niching methods or MMGAs, allow not only to obtain different multiple global optima, but also to preserve useful diversity against convergence to only one solution (the usual behaviour of classical GAs). In previous works, a significant difference in the performance among methods was found, as well as the importance of a suitable parametrization. In this work classic methods are compared to the most recent MMGAs, grouped in three classes (sharing, clearing and species competition), for JSSP. Our experimental study found that those new MMGAs which have a certain type of replacement process perform much better (in terms of highest efficacy and multi-solution based efficacy) than classical MMGAs which do not have this type of process.

Redundant Representations in Evolutionary Computation

This paper discusses how the use of redundant representations influences the performance of genetic and evolutionary algorithms. Representations are redundant if the number of genotypes exceeds the number of phenotypes. A distinction is made between synonymously and non-synonymously redundant representations. Representations are synonymously redundant if the genotypes that represent the same phenotype are very similar to each other. Non-synonymously redundant representations do not allow genetic operators to work properly and result in a lower performance of evolutionary search. When using synonymously redundant representations, the performance of selectorecombinative genetic algorithms (GAs) depends on the modification of the initial supply. We have developed theoretical models for synonymously redundant representations that show the necessary population size to solve a problem and the number of generations goes with O(2(kr)/r), where kr is the order of redundancy and r is the number of genotypic building blocks (BB) that represent the optimal phenotypic BB. As a result, uniformly redundant representations do not change the behavior of GAs. Only by increasing r, which means overrepresenting the optimal solution, does GA performance increase. Therefore, non-uniformly redundant representations can only be used advantageously if a-priori information exists regarding the optimal solution. The validity of the proposed theoretical concepts is illustrated for the binary trivial voting mapping and the real-valued link-biased encoding. Our empirical investigations show that the developed population sizing and time to convergence models allow an accurate prediction of the empirical results.

GA performance distributions and randomly generated binary constraint satisfaction problems

We investigate the variable performance of a genetic algorithm (GA) on randomly generated binary constraint satisfaction problem instances which occur near the phase transition from soluble to non-soluble. We first carry out a conventional landscape analysis and observe, next to a number of common features related to the interaction structure, important differences between the instances, such as the number of solutions, the quality of the paths to the solutions, and the lengths and extent of the neutral paths for mutation. We then split the dynamics of the GA into two phases: the ascent towards the high fitness region, and from this high fitness region to a solution. To gain further information about the GA's behavior in the first phase, we construct two models based on the much simpler fully separable functions, and try to match instances which show a similar performance distribution. Although far from exact, this technique of comparing with analog search problems gives a hint about the landscape elements that are responsible for the GA's slow ascent.

Self-Adaptive Genetic Algorithms with Simulated Binary Crossover

Self-adaptation is an essential feature of natural evolution. However, in the context of function optimization, self-adaptation features of evolutionary search algorithms have been explored mainly with evolution strategy (ES) and evolutionary programming (EP). In this paper, we demonstrate the self-adaptive feature of real-parameter genetic algorithms (GAs) using a simulated binary crossover (SBX) operator and without any mutation operator. The connection between the working of self-adaptive ESs and real-parameter GAs with the SBX operator is also discussed. Thereafter, the self-adaptive behavior of real-parameter GAs is demonstrated on a number of test problems commonly used in the ES literature. The remarkable similarity in the working principle of real-parameter GAs and self-adaptive ESs shown in this study suggests the need for emphasizing further studies on self-adaptive GAs.

Understanding genetic algorithm dynamics using harvesting strategies

Abstract The genetic algorithm (GA) finds optimal solutions over complex fitness landscapes using a method developed in analogy to genetic laws and natural selection. The method essentially operates by optimizing the tradeoff between exploring new points in the search space and exploiting previous information discovered thus far. In this tradeoff, an understanding of the internal GA dynamics, how exactly the GA arrives at an optimum solution, remains somewhat mysterious. Harvesting strategies are introduced here to parameterize the GA's dynamical behavior of elevating sub-threshold solutions toward optimum. The method of harvesting balances the competing aims of population diversity counterweighted against rapid convergence toward the optimum solution. The work establishes that: (1) an upper bound on the fitness ratio exists, above which harvesting becomes too disruptive to the population diversity; (2) analytical conditions for considering elevation within the genetic algorithm are a specific case of logistic growth; and (3) explicit relations exist for the maximum yield and maximum harvestable fraction for 2-stage, 3-stage and finally n-stage harvesting strategies as a function of fitness ratio. Simple expressions for GA time complexity between harvesting steps are presented.

Understanding genetic algorithm dynamics using harvesting strategies

The genetic algorithm (GA) finds optimal solutions over complex fitness landscapes using a method developed in analogy to genetic laws and natural selection. The method essentially operates by optimizing the tradeoff between exploring new points in the search space and exploiting previous information discovered thus far. In this tradeoff, an understanding of the internal GA dynamics, how exactly the GA arrives at an optimum solution, remains somewhat mysterious. Harvesting strategies are introduced here to parameterize the GA's dynamical behavior of elevating sub-threshold solutions toward optimum. The method of harvesting balances the competing aims of population diversity counterweighted against rapid convergence toward the optimum solution. The work establishes that: (1) an upper bound on the fitness ratio exists, above which harvesting becomes too disruptive to the population diversity; (2) analytical conditions for considering elevation within the genetic algorithm are a specific case of logistic growth; and (3) explicit relations exist for the maximum yield and maximum harvestable fraction for 2-stage, 3-stage and finally n-stage harvesting strategies as a function of fitness ratio. Simple expressions for GA time complexity between harvesting steps are presented.

Evolving cellular automata to perform computations: mechanisms and impediments

We present results from experiments in which a genetic algorithm (GA) was used to evolve cellular automata (CAs) to perform a particular computational task - one-dimensional density classification. We look in detail at the evolutionary mechanisms producing the GA's behavior on this task and the impediments faced by the GA. In particular, we identify four “epochs of innovation” in which new CA strategies for solving the problem are discovered by the GA, describe how these strategies are implemented in CA rule tables, and identify the GA mechanisms underlying their discovery. The epochs are characterized by a breaking of the task's symmetries on the part of the GA. The symmetry breaking results in a short-term fitness gain but ultimately prevents the discovery of the most highly fit strategies. We discuss the extent to which symmetry breaking and other impediments are general phenomena in any GA search.