Application Of Evolutionary Computation Research Articles

The application of evolutionary computation in generative adversarial networks (GANs): a systematic literature survey

As a subfield of deep learning (DL), generative adversarial networks (GANs) have produced impressive generative results by applying deep generative models to create synthetic data and by performing an adversarial training process. Nevertheless, numerous issues related to the instability of training need to be urgently addressed. Evolutionary computation (EC), using the corresponding paradigm of biological evolution, overcomes these problems and improves evolutionary-based GANs’ ability to deal with real-world applications. Therefore, this paper presents a systematic literature survey combining EC and GANs. First, the basic theories of GANs and EC are analyzed and summarized. Second, to provide readers with a comprehensive view, this paper outlines the recent advances in combining EC and GANs after detailed classification and introduces each of them. These classifications include evolutionary GANs and their variants, GANs with evolutionary strategies and differential evolution, GANs combined with neuroevolution, evolutionary GANs related to different optimization problems, and applications of evolutionary GANs. Detailed information on the evaluation metrics, network structures, and comparisons of these models is presented in several tables. Finally, future directions and possible perspectives for further development are discussed.

Application of Evolutionary Computation to the Optimization of Biodiesel Mixtures Using a Nature-Inspired Adaptive Genetic Algorithm

The present research work introduces a novel mixture optimization methodology for biodiesel fuels using an Evolutionary Computation method inspired by biological evolution. Specifically, the optimal biodiesel composition is deduced from the application of a nature-inspired adaptive genetic algorithm that first examines percentages of the ingredients in the optimal mixtures. The innovative approach’s effectiveness lies in problem simulation with improvements in the evaluation of the specific function and the way to define and tune the genetic algorithm. Environmental imperatives in the era of climate change currently impose the optimized production of alternative environmentally friendly biofuels to replace fossil fuels. Biodiesel in particular, appears to be more attractive in recent years, as it originates from renewable bio-derived resources. The main ingredients of the specific biofuel mixture investigated in this research are diesel and biodiesel (100% from bioresources). The assessment of the new biodiesel examined was performed using a fitness function that estimated both the density and cost of the fuel. Beyond the evaluation criterion of cost, density also influences the suitability of this biofuel for commercial use and market sale. The outcomes from the modeling process can be beneficial in saving cost and time for new biodiesel production by using this novel decision-making tool in comparison with randomized laboratory experimentations.

Genetic atom search-optimized in vivo bioelectricity harnessing from live dragon fruit plant based on intercellular two-electrode placement

Bioelectricity is a promising alternative renewable energy source that can be produced from live plants and trees. However, previous experimental studies mostly applied non-sustainable bioelectricity extraction techniques from cut-off stem or leaves and neglected the optimum placement of electrodes for maximizing energy extraction without impeding plant growth. Electrode placement and penetration are crucial in energy extraction since they greatly influence electrical generated output enhancement. Relatively, along with the common plants used for bioelectricity extraction, the dragon fruit tree has the potential to be explored as an alternative bioelectricity source since it is widely abundant in many regions. With that, this work introduced a novel integrated genetic-population metaheuristic-based optimization model that was developed centered on in vivo stem bioelectricity extraction from dragon fruit tree to determine the exact optimum distance of silver-coated copper pin-type anodes and cathodes for maximum bioelectricity extraction through intercellular across vascular bundle (icVB) and inter-parenchymal cells (iPC) electrode penetration techniques, and incorporated the cradle-to-gate Life Cycle Assessment methodology to properly account the environmental impacts of the two intercellular penetration approaches. Multigene genetic programming was performed to formulate the fitness function followed by a comparative atom search (ASO), shuffle frog-leaping, and elephant herding-based bioelectricity harnessing optimization. Thus, ASO demonstrated the highest attainable fitness value and conformed well with both electrode placement treatments. This subsequently verified that ASO-based iPC penetration, yielding 58.923 J, surpasses icVB, which only yielded 13.909 J in terms of the total harnessed energy stored throughout the 30-day experiment. Overall, the genetic ASO-iPC with an electrode distance of 4.488 inches produced a higher yield of harnessed bioelectricity while incurring no significant damage and causing fewer environmental impacts compared to the ASO-icVB treatment. This developed technique can minimize greenhouse gas emissions while also expanding the application of evolutionary computing in agriculture and alternative energy domains.

A dynamic trust management model for vehicular ad hoc networks

Trust management in vehicular ad hoc networks (VANETs) is a challenging dynamic optimization problem due to their decentralized, infrastructureless, and dynamically changing topology. Evolutionary computation (EC) algorithms are good candidates for solving dynamic optimization problems (DOPs), since they are inspired from the biological evolution that is occurred as a result of changes in the environment. In this study, we explore the use of genetic programming (GP) algorithm and evolutionary dynamic optimization (EDO) techniques to build a dynamic trust management model for VANETs. The proposed dynamic trust management model properly evaluates the trustworthiness of vehicles and their messages in the simulation of experimental scenarios including bogus information attacks. The simulation results show that the evolved trust calculation formula prevents the propagation of bogus messages over VANETs successfully and the dynamic trust management model detects changes in the problem and reacts to them in a timely manner. The best evolved formula achieves 89.38% Matthews Correlation Coefficient (MCC), 91.81% detection rate (DR), and 1.01% false positive rate (FPR), when ≈ 5% of the network traffic is malicious. The formula obtains 87.33% MCC, 92.01% DR, and 4.8% FPR when ≈ 40% of the network traffic is malicious, demonstrating its robustness to increasing malicious messages. The proposed model is also run on a real-world traffic model and obtains high MCC and low FPR values. To the best of our knowledge, this is the first application of EC and EDO techniques that generate a trust formula automatically for dynamic trust management in VANETs.

Applying Evolutionary Computation to Optimize the Design of Urban Blocks

This empirical study in this paper focuses on the application of evolutionary computation in parametric urban block design to solve the problem of how to effectively find the optimal design solution set among a large number of schemes obtained through parametric urban block design. Through the application of evolutionary computation, a set of parameterized intelligent generation methods of urban blocks under the guidance of multi-conflict objective optimization can be established. The empirical study presented is based on a typical Russian block redevelopment design along the Chinese Eastern Railway. The design aims to transform and redevelop the original block on the basis of protecting its historical buildings and street pattern, taking into account environmental and economic considerations. The final results show that under the premise of reasonable overall evolution, a large number of block design schemes with complex design objectives can be obtained. Furthermore, according to the index data analysis of Pareto’s overall optimal, mean value, median value, extreme value, etc., it provides designers with methods to select a series of optimal schemes quickly and efficiently based on different factors such as comprehensive balance, average and middle ranking, and individual best.

Guest Editorial: Special Issue on Evolutionary Computation for Games

The eight papers in this special section focus on applications of evolutionary computation to games to demonstrate several ways in which evolution can push boundaries and explore new areas of what is possible in the realm of games research, with a focus on game-playing, automatic agent parameter tuning, automatic game testing, and procedural content generation.

Evolutionary Computing to solve product inconsistencies in Software Product Lines

• Inconsistencies prevent and delay the configuration of an SPL product. • Evolutionary computation is proposed to resolve inconsistencies in SPL products. • ga-SPLIS (genetic algorithm-Software Product LIne Solution) is developed. • Results show the applicability of GA to resolve the SPL product inconsistencies. In Software Product Lines (SPLs), multiple design teams work collectively to configure products. Often, having multiple sub-designs leads to inconsistencies which prevent and delay the configuration of the final product at the end. Solutions are available to identify and resolve these inconsistencies for limited scale SPLs. For large scale and industrial SPLs, resolution of inconsistencies is still a potential research topic. In this paper, we propose, implement and validate an application of evolutionary computation, particularly genetic algorithms, to resolve inconsistencies in large scale SPL products. We label our approach ga -SPLIS (genetic algorithm-Software Product LIne Solution). Using a concrete experimental methodology with different feature set sizes and multiple standard performance metrics, we generate results which validate the applicability of ga -SPLIS for the SPL community.

A Review on Convolutional Neural Network Encodings for Neuroevolution

Convolutional neural networks (CNNs) have shown outstanding results in different application tasks. However, the best performance is obtained when customized CNNs architectures are designed, which is labor intensive and requires highly specialized knowledge. Over three decades, neuroevolution (NE) has studied the application of evolutionary computation to optimize artificial neural networks (ANNs) at different levels. It is well known that the encoding of ANNs highly impacts the complexity of the search space and the optimization algorithms&#x2019; performance as well. As NE has rapidly advanced toward the optimization of CNNs topologies, researchers face the challenging duty of representing these complex networks. Furthermore, a compilation of the most widely used encoding methods is nonexistent. In response, we present a comprehensive review on the <i>state-of-the-art</i> of encodings for CNNs.

Classification and Analysis of Optimization Techniques for Integrated Energy Systems Utilizing Renewable Energy Sources: A Review for CHP and CCHP Systems

Energy generation and its utilization is bound to increase in the following years resulting in accelerating depletion of fossil fuels, and consequently, undeniable damages to our environment. Over the past decade, despite significant efforts in renewable energy realization and developments for electricity generation, carbon dioxide emissions have been increasing rapidly. This is due to the fact that there is a need to go beyond the power sector and target energy generation in an integrated manner. In this regard, energy systems integration is a concept that looks into how different energy systems, or forms, can connect together in order to provide value for consumers and producers. Cogeneration and trigeneration are the two most well established technologies that are capable of producing two or three different forms of energy simultaneously within a single system. Integrated energy systems make for a very strong proposition since it results in energy saving, fuel diversification, and supply of cleaner energy. Optimization of such systems can be carried out using several techniques with regards to different objective functions. In this study, a variety of optimization methods that provides the possibility of performance improvements, with or without presence of constraints, are demonstrated, pinpointing the characteristics of each method along with detailed statistical reports. In this context, optimization techniques are classified into two primary groups including unconstrained optimization and constrained optimization techniques. Further, the potential applications of evolutionary computing in optimization of Integrated Energy Systems (IESs), particularly Combined Heat and Power (CHP) and Combined Cooling, Heating, and Power (CCHP), utilizing renewable energy sources are grasped and reviewed thoroughly. It was illustrated that the employment of classical optimization methods is fading out, replacing with evolutionary computing techniques. Amongst modern heuristic algorithms, each method has contributed more to a certain application; while the Genetic Algorithm (GA) was favored for thermoeconomic optimization, Particle Swarm Optimization (PSO) was mostly applied for economic improvements. Given the mathematical nature and constraint satisfaction property of Mixed-Integer Linear Programming (MILP), this method is gaining prominence for scheduling applications in energy systems.

Who Gets Credit for AI-Generated Art?

SummaryThe recent sale of an artificial intelligence (AI)-generated portrait for $432,000 at Christie's art auction has raised questions about how credit and responsibility should be allocated to individuals involved and how the anthropomorphic perception of the AI system contributed to the artwork's success. Here, we identify natural heterogeneity in the extent to which different people perceive AI as anthropomorphic. We find that differences in the perception of AI anthropomorphicity are associated with different allocations of responsibility to the AI system and credit to different stakeholders involved in art production. We then show that perceptions of AI anthropomorphicity can be manipulated by changing the language used to talk about AI—as a tool versus agent—with consequences for artists and AI practitioners. Our findings shed light on what is at stake when we anthropomorphize AI systems and offer an empirical lens to reason about how to allocate credit and responsibility to human stakeholders.

An Experimental Method to Estimate Running Time of Evolutionary Algorithms for Continuous Optimization

Running time analysis is a fundamental problem of critical importance in evolutionary computation. However, the analysis results have rarely been applied to advanced evolutionary algorithms (EAs) in practice, let alone their variants for continuous optimization. In this paper, an experimental method is proposed for analyzing the running time of EAs that are widely used for solving continuous optimization problems. Based on Glivenko–Cantelli theorem, the proposed method simulates the distribution of gain, which is introduced by average gain model to characterize progress during the optimization process. Data fitting techniques are subsequently adopted to obtain a desired function for further analyses. To verify the validity of the proposed method, experiments were conducted to estimate the upper bounds on expected first hitting time of various evolutionary strategies, such as (1, $\lambda $ ) evolution strategy, standard evolution strategy, covariance matrix adaptation evolution strategy, and its improved variants. The results suggest that all estimated upper bounds are correct. Backed up by the proposed method, state-of-the-art EAs for continuous optimization will have identical results about the running time as simplified schemes, which will bridge the gap between theoretical foundation and applications of evolutionary computation.

Urban-Tissue Optimization through Evolutionary Computation

The experiments analyzed in this paper focus their research on the use of Evolutionary Computation (EC) applied to a parametrized urban tissue. Through the application of EC, it is possible to develop a design under a single model that addresses multiple conflicting objectives. The experiments presented are based on Cerdà’s master plan in Barcelona, specifically on the iconic Eixample block which is grouped into a 4 × 4 urban Superblock. The proposal aims to reach the existing high density of the city while reclaiming the block relations proposed by Cerdà’s original plan. Generating and ranking multiple individuals in a population through several generations ensures a flexible solution rather than a single “optimal” one. Final results in the Pareto front show a successful and diverse set of solutions that approximate Cerdà’s and the existing Barcelona’s Eixample states. Further analysis proposes different methodologies and considerations to choose appropriate individuals within the front depending on design requirements.

Guest Editorial Special Issue on Search-Based Software Engineering

It is our pleasure to introduce this Special Issue on Search-Based Software Engineering (SBSE) focusing on the application of evolutionary computation to solve real-world software engineering problems. Evolutionary computation (EC) methods have now become integral part of software engineering. New advancements in EC, such as multi- and many-objective optimization, uncertainty handling for robust and reliable solutions, knowledge discovery and knowledge-augmented EC, dynamic EC, have a great deal of applications in software engineering. Many applications in software engineering have emerged based on the usage of EC for the automation of all phases of the software development process, including the analysis, design, implementation, testing, and maintenance of large software systems. A total of 26 papers were submitted to the Special Issue. Each was subjected to at least three reviews and finally six were accepted for publication as described in the following.

2017 IEEE competition on modern heuristic optimizers for smart grid operation: Testbeds and results

This paper summarizes the two testbeds, datasets, and results of the IEEE PES Working Group on Modern Heuristic Optimization (WGMHO) 2017 Competition on Smart Grid Operation Problems. The competition is organized with the aim of closing the gap between theory and real-world applications of evolutionary computation. Testbed 1 considers stochastic OPF (Optimal Power Flow) based Active-Reactive Power Dispatch (ARPD) under uncertainty and Testbed 2 large-scale optimal scheduling of distributed energy resources. Classical optimization methods are not able to deal with the proposed optimization problems within a reasonable time, often requiring more than one day to provide the optimal solution and a significant amount of memory to perform the computation. The proposed problems can be addressed using modern heuristic optimization approaches, enabling the achievement of good solutions in much lower execution times, adequate for the envisaged decision-making processes. Results from the competition show that metaheuristics can be successfully applied in search of efficient near-optimal solutions for the Stochastic Optimal Power Flow and large-scale energy resource management problems.

Survey on evolutionary computation methods for cybersecurity of mobile ad hoc networks

In this paper, a comprehensive survey of evolutionary computation (EC) methods for cybersecurity of mobile ad hoc networks (MANETs) is presented. Typically, EC methods are named based on the natural processes inspiring them, such as swarm intelligence (e.g., ant colony optimization, artificial bee colony, and particle swarm optimization), evolutionary algorithms (e.g., genetic algorithms, genetic programming, grammatical evolution, and differential evolution), artificial immune systems, and evolutionary games analyzing strategic interactions among different population types. We introduce these methods with their typical applications, and commonly used algorithms to improve cybersecurity within the scope of MANETs. Ongoing and speedy topology changes, multi-hop communication, non-hierarchical organization, and power and computational limitations are among the intrinsic characteristics of MANETs causing cybersecurity vulnerabilities. We describe basic defense mechanisms in MANETs for vulnerability detection, attack deterrence, prevention and recovery, and risk mitigation. We classify principal applications of EC as intrusion detection, trust management, and cryptography in cybersecurity systems to counter measure adversarial activities.

Application of stochastic fractal search in approximation and control of LTI systems

The present work deals with the application of evolutionary computation in approximation and control of linear time invariant (LTI) systems. Stochastic fractal search algorithm (SFS) has been proposed to obtain low order system (LOS) from LTI higher order system (HOS) as well as in speed control of DC motor with PID controller. SFS is quite simple to use in control system and employs the diffusion property present in random fractals to discover the search space. In approximation of LTI systems, the integral square error (ISE) while in control of DC motor, the integral of time multiplied absolute error has been taken as an objective/fitness functions. In system’s approximation, the results show that the proposed SFS based LOS preserves both the transient and steady state properties of original HOS. The simulation results have also been compared in terms of; ISE, integral absolute error and impulse response energy with well known familiar and recently published works in the literature which shows the superiority of SFS algorithm. In control of DC motor, the obtained results are satisfactory having no overshoot and less rise and settling times in comparison to existing techniques.

DICE: exploiting all bivariate dependencies in binary and multary search spaces

Although some of the earliest Estimation of Distribution Algorithms (EDAs) utilized bivariate marginal distribution models, up to now, all discrete bivariate EDAs had one serious limitation: they were constrained to exploiting only a limited O(d) subset out of all possible $$O(d^{2})$$ bivariate dependencies. As a first we present a family of discrete bivariate EDAs that can learn and exploit all $$O(d^{2})$$ dependencies between variables, and yet have the same run-time complexity as their more limited counterparts. This family of algorithms, which we label DICE (DIscrete Correlated Estimation of distribution algorithms), is rigorously based on sound statistical principles, and particularly on a modelling technique from statistical physics: dichotomised multivariate Gaussian distributions. Initially (Lane et al. in European Conference on the Applications of Evolutionary Computation, Springer, 1999), DICE was trialled on a suite of combinatorial optimization problems over binary search spaces. Our proposed dichotomised Gaussian (DG) model in DICE significantly outperformed existing discrete bivariate EDAs; crucially, the performance gap increasingly widened as dimensionality of the problems increased. In this comprehensive treatment, we generalise DICE by successfully extending it to multary search spaces that also allow for categorical variables. Because correlation is not wholly meaningful for categorical variables, interactions between such variables cannot be fully modelled by correlation-based approaches such as in the original formulation of DICE. Therefore, here we extend our original DG model to deal with such situations. We test DICE on a challenging test suite of combinatorial optimization problems, which are defined mostly on multary search spaces. While the two versions of DICE outperform each other on different problem instances, they both outperform all the state-of-the-art bivariate EDAs on almost all of the problem instances. This further illustrates that these innovative DICE methods constitute a significant step change in the domain of discrete bivariate EDAs.

Application of evolutionary computation on ensemble forecast of quantitative precipitation

An evolutionary computation algorithm known as genetic programming (GP) has been explored as an alternative tool for improving the ensemble forecast of 24-h accumulated precipitation. Three GP versions and six ensembles’ languages were applied to several real-world datasets over southern, southeastern and central Brazil during the rainy period from October to February of 2008–2013. According to the results, the GP algorithms performed better than two traditional statistical techniques, with errors 27–57% lower than simple ensemble mean and the MASTER super model ensemble system. In addition, the results revealed that GP algorithms outperformed the best individual forecasts, reaching an improvement of 34–42%. On the other hand, the GP algorithms had a similar performance with respect to each other and to the Bayesian model averaging, but the former are far more versatile techniques. Although the results for the six ensembles’ languages are almost indistinguishable, our most complex linear language turned out to be the best overall proposal. Moreover, some meteorological attributes, including the weather patterns over Brazil, seem to play an important role in the prediction of daily rainfall amount.

Evolutionary computing applied to customer relationship management: A survey

Customer relationship management (CRM) is a customer-centric business strategy which a company employs to improve customer experience and satisfaction by customizing products and services to customers' needs. This strategy, when implemented in totality eventually increases the revenue of the company. Traditionally, data mining (DM) techniques have been applied to solve various analytical CRM tasks. In turn, optimization techniques have long been used for training some of the DM techniques. However, during the past few years, evolutionary techniques have become so powerful and versatile that they can be deployed as a substitute for some DM techniques. This trend caught the attention of the researchers working in the analytical CRM area as they too started solving the CRM tasks using evolutionary techniques alone. In this context, we present a survey of evolutionary computing techniques applied to CRM tasks. In this paper, we surveyed 78 papers that were published during 1998 and 2015, where the application of evolutionary computing (EC) techniques to analytical CRM tasks is the main focus. The survey includes papers involving evolutionary computing techniques applied to the analytical CRM tasks under single- as well as multi-objective optimization framework. The purpose of the survey is to let the reader realize the versatility and power of EC techniques in solving analytical CRM tasks in the service industry and suggesting future directions.

A Comprehensive Review on the Application of Evolutionary Computation in Design Optimization of Plate-Fin Heat Exchangers

This review study presents the utilization of different evolutionary algorithms (EAs) for plate-fin heat exchanger (PFHE) design optimization in the last decade. The traditional trial-and-error design process of CHEs can be replaced by a more flexible and easy-to-implement evolutionary approach. EAs do need any information of derivatives and therefore can be implemented in a wide range of applications without much effort. Many EAs have been successfully presented and employed in various fields of engineering including heat transfer problems and more specifically CHE design optimization. However, there seems to be gap in the literature regarding analyses of pros and cons of these algorithms for use in this field. This study could be useful for future use of EAs in not only CHE design optimization, but also for the whole field of heat exchanger.