Performance Of Evolutionary Algorithms Research Articles

An analysis of the interface between evolutionary algorithm operators and problem features for water resources problems. A case study in water distribution network design

Evolutionary Algorithms (EAs) have been widely employed to solve water resources problems for nearly two decades with much success. However, recent research in hyperheuristics has raised the possibility of developing optimisers that adapt to the characteristics of the problem being solved. In order to select appropriate operators for such optimisers it is necessary to first understand the interaction between operator and problem. This paper explores the concept of EA operator behaviour in real world applications through the empirical study of performance using water distribution networks (WDN) as a case study. Artificial networks are created to embody specific WDN features which are then used to evaluate the impact of network features on operator performance. The method extracts key attributes of the problem which are encapsulated in the natural features of a WDN, such as topologies and assets, on which different EA operators can be tested. The method is demonstrated using small exemplar networks designed specifically so that they isolate individual features. A set of operators are tested on these artificial networks and their behaviour characterised. This process provides a systematic and quantitative approach to establishing detailed information about an algorithm's suitability to optimise certain types of problem. The experiment is then repeated on real-world inspired networks and the results are shown to fit with the expected results.

HYBRID NSGA-II OPTIMIZATION FOR IMPROVING THE THREE-TERM BP NETWORK FOR MULTICLASS CLASSIFICATION PROBLEMS

Recently, hybrid algorithms have received considerable attention from a number of researchers. This paper presents a hybrid of the multiobjective evolutionary algorithm to gain a better accuracy of the fi nal solutions. The aim of using the hybrid algorithm is to improve the multiobjective evolutionary algorithm performance in terms of the enhancement of all the individuals in the population and increase the quality of the Pareto optimal solutions. The multiobjective evolutionary algorithm used in this study is a nondominated sorting genetic algorithm-II (NSGA-II) together with its hybrid, the backpropagation algorithm (BP), which is used as a local search algorithm to optimize the accuracy and complexity of the three-term backpropagation (TBP) network. The outcome positively demonstrates that the hybrid algorithm is able to improve the classification performance with a smaller number of hidden nodes and is effective in multiclass classifi cation problems. Furthermore, the results indicate that the proposed hybrid method is a potentially useful classifi er for enhancing the classification process ability when compared with the multiobjective genetic algorithm based on the TBP network (MOGATBP) and certain other methods found in the literature.

A directed search strategy for evolutionary dynamic multiobjective optimization

Many real-world multiobjective optimization problems are dynamic, requiring an optimization algorithm that is able to continuously track the moving Pareto front over time. In this paper, we propose a directed search strategy (DSS) consisting of two mechanisms for improving the performance of multiobjective evolutionary algorithms in changing environments. The first mechanism reinitializes the population based on the predicted moving direction as well as the directions that are orthogonal to the moving direction of the Pareto set, when a change is detected. The second mechanism aims to accelerate the convergence by generating solutions in predicted regions of the Pareto set according to the moving direction of the non-dominated solutions between two consecutive generations. The two mechanisms, when combined together, are able to achieve a good balance between exploration and exploitation for evolutionary algorithms to solve dynamic multiobjective optimization problems. We compare DSS with two existing prediction strategies on a variety of test instances having different changing dynamics. Empirical results show that DSS is powerful for evolutionary algorithms to deal with dynamic multiobjective optimization problems.

Effectiveness of evolutionary algorithms for optimization of heat exchangers

This paper comprehensively investigates performance of evolutionary algorithms for design optimization of shell and tube heat exchangers (STHX). Genetic algorithm (GA), firefly algorithm (FA), and cuckoo search (CS) method are implemented for finding the optimal values for seven key design variables of the STHX model. ∊-NTU method and Bell-Delaware procedure are used for thermal modeling of STHX and calculation of shell side heat transfer coefficient and pressure drop. The purpose of STHX optimization is to maximize its thermal efficiency. Obtained results for several simulation optimizations indicate that GA is unable to find permissible and optimal solutions in the majority of cases. In contrast, design variables found by FA and CS always lead to maximum STHX efficiency. Also computational requirements of CS method are significantly less than FA method. As per optimization results, maximum efficiency (83.8%) can be achieved using several design configurations. However, these designs are bearing different dollar costs. Also it is found that the behavior of the majority of decision variables remains consistent in different runs of the FA and CS optimization processes.

A hybrid evolutionary multiobjective optimization algorithm with adaptive multi-fitness assignment

There are several studies on hybrid multi-operator recombination methods, while few works have been proposed in the area of combining different fitness assignment in a framework. On the other hand, it is known that fitness assignment has a marked impact on the performance of evolutionary multiobjective optimization algorithm (EMOA). In this paper, a hybrid EMOA is proposed, which divides the population into several smaller subpopulations according to their distribution in the objective space. Each subpopulation is evolved by an individual EMOA, and a hybrid performance measure estimates the performance of these EMOAs. We focus on the fitness assignment and assume that all EMOAs used in the subpopulations adopt the same recombination operator. To evaluate performance of the proposed algorithm, we compare it with MOEA/D-M2M, MOE-A/D, SMS-EMOA and NSGA-II on 16 test instances. Experimental results show that the proposed algorithm performs better than or similar to those compared EMOAs.

The influences of canonical evolutionary algorithm operators and variable orderings in learning Bayesian classifiers from data

Variable Ordering (VO) plays an important role when inducing Bayesian Networks (BNs) and Bayesian Classifiers (BCs). Previous works in the literature suggest that it is worth pursuing the use of genetic/evolutionary algorithms for identifying a suitable VO, when learning a BN structure from data. This paper proposes a collaborative Evolutionary-Bayes algorithm named VOEA (Variable Ordering Evolutionary Algorithm) aimed at inducing BCs from data. The two VOEA versions presented in the paper refine a previously proposed algorithm named VOGA by employing only a single evolutionary operator (either crossover or mutation) as well as by using information about the class variable when defining the most suitable variable ordering for learning a BC. Experiments performed in a number of datasets revealed that the VOEA approach is promising and tends to generate suitable and representative BCs, particularly in its version VOEA_M, which only implements the mutation operator.

Analyzing convergence performance of evolutionary algorithms: A statistical approach

The analysis of the performance of different approaches is a staple concern in the design of Computational Intelligence experiments. Any proper analysis of evolutionary optimization algorithms should incorporate a full set of benchmark problems and state-of-the-art comparison algorithms. For the sake of rigor, such an analysis may be completed with the use of statistical procedures, supporting the conclusions drawn.In this paper, we point out that these conclusions are usually limited to the final results, whereas intermediate results are seldom considered. We propose a new methodology for comparing evolutionary algorithms’ convergence capabilities, based on the use of Page’s trend test. The methodology is presented with a case of use, incorporating real results from selected techniques of a recent special issue. The possible applications of the method are highlighted, particularly in those cases in which the final results do not enable a clear evaluation of the differences among several evolutionary techniques.

Optimised Random Mutations for Evolutionary Algorithms

To demonstrate our approaches we will use Sudoku puzzles, which are an excellent test bed for evolutionary algorithms. The puzzles are accessible enough for people to enjoy. However the more complex puzzles require thousands of iterations before an evolutionary algorithm finds a solution. If we were attempting to compare evolutionary algorithms we could count their iterations to solution as an indicator of relative efficiency. Evolutionary algorithms however include a process of random mutation for solution candidates. We will show that by improving the random mutation behaviours we were able to solve problems with minimal evolutionary optimisation. Experiments demonstrated the random mutation was at times more effective at solving the harder problems than the evolutionary algorithms. This implies that the quality of random mutation may have a significant impact on the performance of evolutionary algorithms with Sudoku puzzles. Additionally this random mutation may hold promise for reuse in hybrid evolutionary algorithm behaviours.

Biogeography-based optimisation with chaos

The biogeography-based optimisation (BBO) algorithm is a novel evolutionary algorithm inspired by biogeography. Similarly, to other evolutionary algorithms, entrapment in local optima and slow convergence speed are two probable problems it encounters in solving challenging real problems. Due to the novelty of this algorithm, however, there is little in the literature regarding alleviating these two problems. Chaotic maps are one of the best methods to improve the performance of evolutionary algorithms in terms of both local optima avoidance and convergence speed. In this study, we utilise ten chaotic maps to enhance the performance of the BBO algorithm. The chaotic maps are employed to define selection, emigration, and mutation probabilities. The proposed chaotic BBO algorithms are benchmarked on ten test functions. The results demonstrate that the chaotic maps (especially Gauss/mouse map) are able to significantly boost the performance of BBO. In addition, the results show that the combination of chaotic selection and emigration operators results in the highest performance.

Multiobjective memetic algorithm based on decomposition

In recent years, hybridization of multi-objective evolutionary algorithms (MOEAs) with traditional mathematical programming techniques have received significant attention in the field of evolutionary computing (EC). The use of multiple strategies with self-adaptation manners can further improve the algorithmic performances of decomposition-based evolutionary algorithms. In this paper, we propose a new multiobjective memetic algorithm based on the decomposition approach and the particle swarm optimization (PSO) algorithm. For brevity, we refer to our developed approach as MOEA/D-DE+PSO. In our proposed methodology, PSO acts as a local search engine and differential evolution works as the main search operator in the whole process of optimization. PSO updates the position of its solution with the help of the best information on itself and its neighboring solution. The experimental results produced by our developed memtic algorithm are more promising than those of the simple MOEA/D algorithm, on most test problems. Results on the sensitivity of the suggested algorithm to key parameters such as population size, neighborhood size and maximum number of solutions to be altered for a given subproblem in the decomposition process are also included.

Hypervolume Performance of Conical Area Evolutionary Algorithm for Bi-Objective Optimization

The conical area evolutionary algorithm (CAEA) can efficiently solve the bi-objective optimization problems by borrowing some ideas from decomposition and hypervolume. In this paper, the optimal hypervolume performance of the CAEA with an infinite number of sub-problems is proved through the squeeze theorem for limits. Experimental results on several bi-objective optimization problems have shown that not only CAEA performs much better than NSGA-II and MOEA/D in terms of efficiency, but also CAEA with a larger number of sub-problems has the better hypervolume performance.

Factorial design analysis applied to the performance of parallel evolutionary algorithms

BackgroundParallel computing is a powerful way to reduce computation time and to improve the quality of solutions of evolutionary algorithms (EAs). At first, parallel EAs (PEAs) ran on very expensive and not easily available parallel machines. As multicore processors become ubiquitous, the improved performance available to parallel programs is a great motivation to computationally demanding EAs to turn into parallel programs and exploit the power of multicores. The parallel implementation brings more factors to influence performance and consequently adds more complexity on PEA evaluations. Statistics can help in this task and can guarantee the significance and correct conclusions with minimum tests, provided that the correct design of experiments is applied.MethodsWe show how to guarantee the correct estimation of speedups and how to apply a factorial design on the analysis of PEA performance.ResultsThe performance and the factor effects were not the same for the two benchmark functions studied in this work. The Rastrigin function presented a higher coefficient of variation than the Rosenbrock function, and the factor and interaction effects on the speedup of the parallel genetic algorithm I (PGA-I) were different in both.ConclusionsAs a case study, we evaluate the influence of migration related to parameters on the performance of the parallel evolutionary algorithm solving two benchmark problems executed on a multicore processor. We made a particular effort in carefully applying the statistical concepts in the development of our analysis.

Performance Metric Ensemble for Multiobjective Evolutionary Algorithms

Evolutionary algorithms have been successfully exploited to solve multiobjective optimization problems. In the literature, a heuristic approach is often taken. For a chosen benchmark problem with specific problem characteristics, the performance of multiobjective evolutionary algorithms (MOEAs) is evaluated via some heuristic chosen performance metrics. The conclusion is then drawn based on statistical findings given the preferable choices of performance metrics. The conclusion, if any, is often indecisive and reveals no insight pertaining to which specific problem characteristics the underlying MOEA could perform the best. In this paper, we introduce an ensemble method to compare MOEAs by combining a number of performance metrics using double elimination tournament selection. The double elimination design allows characteristically poor performance of a quality algorithm to still be able to win it all. Experimental results show that the proposed metric ensemble can provide a more comprehensive comparison among various MOEAs than what could be obtained from a single performance metric alone. The end result is a ranking order among all chosen MOEAs, but not quantifiable measures pertaining to the underlying MOEAs.

Automatic parameter tuning for Evolutionary Algorithms using a Bayesian Case-Based Reasoning system

The widespread use and applicability of Evolutionary Algorithms is due in part to the ability to adapt them to a particular problem-solving context by tuning their parameters. This is one of the problems that a user faces when applying an Evolutionary Algorithm to solve a given problem. Before running the algorithm, the user typically has to specify values for a number of parameters, such as population size, selection rate, and probability operators.This paper empirically assesses the performance of an automatic parameter tuning system in order to avoid the problems of time requirements and the interaction of parameters. The system, based on Bayesian Networks and Case-Based Reasoning methodology, estimates the best parameter setting for maximizing the performance of Evolutionary Algorithms. The algorithms are applied to solve a basic problem in constraint-based, geometric parametric modeling, as an instance of general constraint-satisfaction problems.The experimental results demonstrate the validity of the proposed system and its potential effectiveness for configuring algorithms.

Designing a parallel evolutionary algorithm for inferring gene networks on the cloud computing environment

BackgroundTo improve the tedious task of reconstructing gene networks through testing experimentally the possible interactions between genes, it becomes a trend to adopt the automated reverse engineering procedure instead. Some evolutionary algorithms have been suggested for deriving network parameters. However, to infer large networks by the evolutionary algorithm, it is necessary to address two important issues: premature convergence and high computational cost. To tackle the former problem and to enhance the performance of traditional evolutionary algorithms, it is advisable to use parallel model evolutionary algorithms. To overcome the latter and to speed up the computation, it is advocated to adopt the mechanism of cloud computing as a promising solution: most popular is the method of MapReduce programming model, a fault-tolerant framework to implement parallel algorithms for inferring large gene networks.ResultsThis work presents a practical framework to infer large gene networks, by developing and parallelizing a hybrid GA-PSO optimization method. Our parallel method is extended to work with the Hadoop MapReduce programming model and is executed in different cloud computing environments. To evaluate the proposed approach, we use a well-known open-source software GeneNetWeaver to create several yeast S. cerevisiae sub-networks and use them to produce gene profiles. Experiments have been conducted and the results have been analyzed. They show that our parallel approach can be successfully used to infer networks with desired behaviors and the computation time can be largely reduced.ConclusionsParallel population-based algorithms can effectively determine network parameters and they perform better than the widely-used sequential algorithms in gene network inference. These parallel algorithms can be distributed to the cloud computing environment to speed up the computation. By coupling the parallel model population-based optimization method and the parallel computational framework, high quality solutions can be obtained within relatively short time. This integrated approach is a promising way for inferring large networks.

The Effects of Using Chaotic Map on Improving the Performance of Multiobjective Evolutionary Algorithms

Chaotic maps play an important role in improving evolutionary algorithms (EAs) for avoiding the local optima and speeding up the convergence. However, different chaotic maps in different phases have different effects on EAs. This paper focuses on exploring the effects of chaotic maps and giving comprehensive guidance for improving multiobjective evolutionary algorithms (MOEAs) by series of experiments. NSGA-II algorithm, a representative of MOEAs using the nondominated sorting and elitist strategy, is taken as the framework to study the effect of chaotic maps. Ten chaotic maps are applied in MOEAs in three phases, that is, initial population, crossover, and mutation operator. Multiobjective problems (MOPs) adopted are ZDT series problems to show the generality. Since the scale of some sequences generated by chaotic maps is changed to fit for MOPs, the correctness of scaling transformation of chaotic sequences is proved by measuring the largest Lyapunov exponent. The convergence metricγand diversity metric Δ are chosen to evaluate the performance of new algorithms with chaos. The results of experiments demonstrate that chaotic maps can improve the performance of MOEAs, especially in solving problems with convex and piecewise Pareto front. In addition, cat map has the best performance in solving problems with local optima.

Evolutionary modeling for parameter estimation for chaotic system

On the basis of evolutionary algorithm, a novel method for parameter estimation of nonlinear dynamic equations is given in the present paper. Numerical tests indicate that the unknown parameters all can be estimated quickly and accurately whether the partial parameters are unknown or all parameters are unknown in the classic Lorenz equation. However, it is found that the convergence rate of the new algorithm is relatively slow when multiple unknown parameters are estimated simultaneously. To solve this problem, a corresponding improvement of measure is proposed, namely, a constraint mechanism is taken during the variation operation of evolutionary algorithm. The improvement is mainly based on the characteristic that the longer the running time of the evolutionary algorithm, the smaller the range of variation of the estimated parameters. Results indicate that the searching speed of the algorithm is greatly improved by using the improved estimation parameter project.

A Mathematical Design of Genetic Operators onGLn(ℤ2)

We study the space that consists of all nonsingular binary matrices, that is,GLn(ℤ2). The space is quite important in that it is used for the change of basis in binary representation, which is the encoding typically adopted in genetic algorithms. We analyze the properties ofGLn(ℤ2)and theoretically design possible encodings and their corresponding recombination operators for evolutionary algorithms. We present approaches based on elementary matrices of linear algebra as well as typical two-dimensional ones.

Transgenic: An evolutionary algorithm operator

Traditionally, many evolutionary algorithm operators have biological inspiration. Genetics has contributed to the proposal of a number of different evolutionary operators, such as haploid crossover, mutation, diploid, inversion, gene doubling, deletion, and others. In the present study, we propose a new genetic-inspired evolutionary operator, named Transgenic, which was specially designed for Genetic Algorithms (GA). The proposed operator is inspired by genetically modified organisms (GMOs), where important features are artificially introduced into their genome. Transgenic can be used to artificially insert relevant characteristics in the chromosome of individuals, thus converging to better results faster than traditional GAs. When relevant characteristics are known a prior, then, Transgenic simply forces the presence of such characteristics in part of the population (in an elitism-based approach). Whenever there is no a priori knowledge available, Transgenic automatically identifies relevant features (based on historical information) to perform the elitism approach. The GA, used in this study was designed to allow the discovery of concise, yet accurate, high-level rules (from synthetic and real biological databases) which can be used as a classification system. The empirical results have shown that Transgenic is capable of generating better results than traditional rule classification methods, such as J48, Single Conjunctive Rule Learner, One R and PART, using synthetic datasets.

Prediction in evolutionary algorithms for dynamic environments

Evolutionary algorithms have been widely used to solve dynamic optimization problems. Memory-based evolutionary algorithms are often used when the dynamics of the environment follow some repeated behavior. Over the last few years, the use of prediction mechanisms combined with memory has been explored. These prediction techniques are used to avoid the decrease of the algorithm's performance when a change occurs. This paper investigates the use of prediction methods in memory-based evolutionary algorithms for two distinct situations: to predict when the next change will happen and how the environment will change. For the first predictor two techniques are explored, one based on linear regression and another supported by nonlinear regression. For the second, a technique based on Markov chains is explored. Several experiments were carried out using different types of dynamics in two benchmark problems. Experimental results show that the incorporation of the proposed prediction techniques efficiently improves the performance of evolutionary algorithms in dynamic optimization problems.