Evolutionary Algorithm Approach Research Articles

Integrated neural network and evolutionary algorithm approach for liver fibrosis staging: Can artificial intelligence reduce patient costs?

Staging liver fibrosis is important, and liver biopsy is the gold standard diagnostic tool. We aim to design and evaluate an artificial neural network (ANN) method by taking advantage of the Teaching Learning-Based Optimization (TLBO) algorithm for the prediction of liver fibrosis stage in blood donors and hepatitis C patients. We propose a method based on a selection of machine learning classification methods including multilayer perceptron (MLP) neural network, Naive Bayesian (NB), decision tree, and deep learning. Initially, the synthetic minority oversampling technique (SMOTE) is performed to address the imbalance in the dataset. Afterward, the integration of MLP and TLBO is implemented. We propose a novel algorithm that reduces the number of required patient features to seven inputs. The accuracy of MLP using 12 features is 0.903, while that of the proposed MLP with TLBO is 0.891. Besides, the diagnostic accuracy of all methods, except the model designed with the Bayesian network, increases when the SMOTE balancer is applied. The decision tree-based deep learning methods show the highest levels of accuracy with 12 features. Interestingly, with the use of TLBO and seven features, MLP reached an accuracy rate of 0.891, which is quite satisfactory when compared with those of similar studies. The proposed model provides high diagnostic accuracy, while reducing the required number of properties from the samples. The results of our study show that the recruited algorithm of our study is more straightforward, with a smaller number of required properties and similar accuracy.

AxOMaP : Designing FPGA-based A ppro x imate Arithmetic O perators using Ma thematical P rogramming

With the increasing application of machine learning (ML) algorithms in embedded systems, there is a rising necessity to design low-cost computer arithmetic for these resource-constrained systems. As a result, emerging models of computation, such as approximate and stochastic computing, that leverage the inherent error-resilience of such algorithms are being actively explored for implementing ML inference on resource-constrained systems. Approximate computing (AxC) aims to provide disproportionate gains in the power, performance, and area (PPA) of an application by allowing some level of reduction in its behavioral accuracy (BEHAV). Using approximate operators (AxOs) for computer arithmetic forms one of the more prevalent methods of implementing AxC. AxOs provide the additional scope for finer granularity of optimization, compared to only precision scaling of computer arithmetic. To this end, the design of platform-specific and cost-efficient approximate operators forms an important research goal. Recently, multiple works have reported the use of AI/ML-based approaches for synthesizing novel FPGA-based AxOs. However, most of such works limit the use of AI/ML to designing ML-based surrogate functions that are used during iterative optimization processes. To this end, we propose a novel data analysis-driven mathematical programming-based approach to synthesizing approximate operators for FPGAs. Specifically, we formulate mixed integer quadratically constrained programs based on the results of correlation analysis of the characterization data and use the solutions to enable a more directed search approach for evolutionary optimization algorithms. Compared to traditional evolutionary algorithms-based optimization, we report up to 21% improvement in the hypervolume, for joint optimization of PPA and BEHAV, in the design of signed 8-bit multipliers. Further, we report up to 27% better hypervolume than other state-of-the-art approaches to DSE for FPGA-based application-specific AxOs.

A Comparative Study of Evolutionary and Swarm Intelligence Algorithms for Job Scheduling on Identical Parallel Machines

In parallel computing systems, job scheduling plays a crucial role in enhancing system efficiency and minimizing the makespan. In recent years, evolutionary and swarm intelligence algorithms have gained prominence as effective approaches for solving combinatorial optimization problems. In the present work, we have considered genetic algorithm (GA) for evolutionary algorithms and particle swarm optimization (PSO) for swarm intelligence algorithms. Evolutionary algorithms (EA) and swarm intelligence algorithms (SIA) have shown promising results in solving job scheduling challenges. In this study, we collate the performance of EA and SIA approaches for job scheduling on parallel machines. We use different benchmark instances to evaluate the algorithms' makespan and computational time performance. The results show that SIA algorithms outperform EA algorithms regarding makespan and computational time for all benchmark instances. Furthermore, the study provides insights into the strengths and weaknesses of EA and SIA algorithms for job scheduling on parallel machines. Our findings provide useful insights for researchers and practitioners interested in applying optimization techniques to solve job scheduling problems on parallel machines.

Evolving the MCTS Upper Confidence Bounds for Trees Using a Semantic-Inspired Evolutionary Algorithm in the Game of Carcassonne

Monte Carlo Tree Search (MCTS) is a sampling best-first method to search for optimal decisions. The success of MCTS depends heavily on how the tree is built and the selection process plays a fundamental role in this. One particular selection mechanism that has proved to be reliable is based on the Upper Confidence Bounds for Trees (UCT). The UCT attempts to balance exploration and exploitation by considering the values stored in the statistical tree of the MCTS. However, some tuning of the MCTS UCT is necessary for this to work well. In this work, we use Evolutionary Algorithms (EAs) to evolve mathematical expressions with the goal to substitute the UCT formula and use the evolved expressions in MCTS. More specifically, we evolve expressions by means of our proposed Semantic-inspired Evolutionary Algorithm in MCTS approach (SIEA-MCTS). This is inspired by semantics in Genetic Programming (GP), where the use of fitness cases is seen as a requirement to be adopted in GP. Fitness cases are normally used to determine the fitness of individuals and can be used to compute the semantic similarity (or dissimilarity) of individuals. However, fitness cases are not available in MCTS. We extend this notion by using multiple reward values from MCTS that allow us to determine both the fitness of an individual and its semantics. By doing so, we show how SIEA-MCTS is able to successfully evolve mathematical expressions that yield better or competitive results compared to UCT without the need of tuning these evolved expressions. We compare the performance of the proposed SIEA-MCTS against MCTS algorithms, MCTS Rapid Action Value Estimation algorithms, three variants of the *-minimax family of algorithms, a random controller and two more EA approaches. We consistently show how SIEA-MCTS outperforms most of these intelligent controllers in the challenging game of Carcassonne, whose state-space complexity is, approx., 10 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^{40}$</tex-math></inline-formula> .

A hybrid multi-start metaheuristic scheduler for astronomical observations

In this paper, we investigate Astronomical Observations Scheduling which is a type of Multi-Objective Combinatorial Optimization Problem, and detail its specific challenges and requirements and propose the Hybrid Accumulative Planner (HAP), a hybrid multi-start metaheuristic scheduler able to adapt to the different variations and demands of the problem. To illustrate the capabilities of the proposal in a real-world scenario, HAP is tested on the Atmospheric Remote-sensing Infrared Exoplanet Large-survey (Ariel) mission of the European Space Agency (ESA), and compared with other studies on this subject including an Evolutionary Algorithm (EA) approach. The results show that the proposal outperforms the other methods in the evaluation and achieves better scientific goals than its peers. The consistency of HAP in obtaining better results on the available datasets for Ariel, with various sizes and constraints, demonstrates its competence in scalability and adaptability to different conditions of the problem.

High-pressure structure prediction and physical properties of MnN2: First-principles calculations

Manganese nitrides have attracted considerable research interest since their rich physical properties and important application prospects. By employing an evolutionary algorithm approach and first-principles calculations, the stable crystal structures of MnN2 in the pressure range of 0–200 GPa are systematically explored. Two novel high-pressure MnN2 phases, Pm and Cmc21, are predicted. The Pm phase is the most stable structure within the pressure range of 153.5–155.9 GPa, and it transforms into Cmc21 configuration at 155.9 GPa. By calculating the phonon spectra and molecular dynamics simulations, the dynamical stability and thermodynamic stability are examined. To uncover the potential applications of them, we investigate their mechanical properties and magnetism. Pm-MnN2, as a metal system, exhibits outstanding bulk modulus (296.1 GPa) and impressively high Vickers hardness (up to 30.5 GPa). Furthermore, the electronic mechanism underlying the excellent hardness of Pm-MnN2 is discussed by analyzing the crystal orbital Hamilton populations. Besides, Cmc21-MnN2 exhibits ferromagnetic configurations.

Bi-EvoGAN: Bi-level Evolutionary Approach for Generative Adversarial Networks

Generative Adversarial Networks (GANs) are an adversarial approach to generative modeling using deep learning methods, which have become one of the most relevant topics in the last years. Despite producing interesting outputs, recent studies have demonstrated that GANs remain difficult to control and suffer from training problems, such as instability and mode collapse, as well as several aspects of their models are designed by hand. Neuroevolution is a recent evolution strategy that uses Evolutionary Algorithms (EAs) in order to provide an automatic design of deep neural networks architectures. Many recent efforts have been dedicated to contributing on GANs by developing different single-level EAs approaches, however they could not solve their common issues completely which are still open problems. In this paper, we propose a new design of evolutionary algorithms in the context of GANs, consisting of dividing the evolution search space into two complementary levels, modeling a Bi-level Evolutionary Approach called Bi-EvoGAN searching to surpass their adversarial optimization difficulties and also to demonstrate the performance of considering two levels of optimization in comparison to a single optimization level. To the best of our knowledge, this work presents the first application of the bi-level optimization in the context of generative adversarial networks. The upper level uses an Evolutionary Algorithm-based Topology Optimization method named “EA-TO” to evolve a population of GAN networks as a set of individuals in order to find the best network topology. The lower level applies an Evolutionary Algorithm-based Decomposition for Hyperparameter Optimization termed “EA/D-HO” in order to select the best hyperparameters of each GAN individual. The performance of Bi-EvoGAN is explained by the double optimization of the GAN topology and their hyperparameters through the new bilevel optimization design allowing to evolve a GAN population in its higher fitness which will improve the quality of the search method and consequently to accelerate the convergence. Experimental studies are carried out based on four well-known benchmark datasets, MNIST, Fashion-MNIST, CelebA and LSUN-Bedroom allowing to show the efficiency of our approach in comparison to the other state-of-the-art approaches.

A Novel EA-Based Techno–Economic Analysis of Charging System for Electric Vehicles: A Case Study of Qassim Region, Saudi Arabia

Because of the fast expansion of electric vehicles (EVs) in Saudi Arabia, a massive amount of energy will be needed to serve these vehicles. In addition, the transportation sector radiates a considerable amount of toxic gases in the form of SO2 and CO2. The national grid must supply a huge amount of electricity on a regular basis to meet the increasing power demands of EVs. This study thoroughly investigates the technical and economic benefits of an off-grid and grid-connected hybrid energy system with various configurations of a solar, wind turbine and battery energy storage system for the electric vehicle charging load in the Qassim region, Saudi Arabia. The goal is to decrease the cost of energy while reducing the chance of power outages in the system. This is achieved by using a new optimization algorithm called the modified salp swarm optimization algorithm (MSSOA), which is based on an evolutionary algorithm approach. MSSOA is an improved version of SSOA, which addresses its shortcomings. It has two search strategies to enhance its efficiency: first, it uses Levy flight distribution (LFD) to help individuals reach new positions faster, and second, it instructs individuals to spiral around the optimal solution, improving the exploitation phase. The MSSOA’s effectiveness is confirmed by comparing its results with those of the conventional salp swarm optimization algorithm and particle swarm optimization (PSO). According to simulation findings, MSSOA has excellent accuracy and robustness. In this region, the SPV/WT/BESS-based EV charging station is the optimal option for EV charging stations. The SPV/WT/BESS design has the lowest LCOE of all feasible configurations in the region under study. The optimum values for the LCOE and TNPC using MSSOA are USD 0.3697/kWh and USD 99,928.34, which are much lower than the optimized values for the LCOE (USD 0.4156) and TNPC (USD 1,12,671.75) using SSOA. Furthermore, a comprehensive techno–economic analysis of optimized hybrid systems is assessed by incorporating the grid-connected option. The grid connected system results in optimized values of the LCOE (USD 0.0732/kWh) and TNPC (USD 1,541,076). The impact of different grid purchase prices on the levelized cost of energy is also studied. Our results will assist the researchers to determine the best technique for the optimization of an optimal energy system.

Pressure-induced novel ZrN4 semiconductor materials with high dielectric constants: a first-principles study.

In addition to Zr3N4 and ZrN2 compounds, zirconium nitrides with a rich family of phases always exhibit metal phases. By employing an evolutionary algorithm approach and first-principles calculations, we predicted seven novel semiconductor phases for the ZrN4 system at 0-150 GPa. Through calculating phonon dispersions, we identified four dynamically stable semiconductor structures under ambient pressure, namely, α-P1̄, β-P1̄, γ-P1̄, and β-P1 (with bandgaps of 1.03 eV, 1.10 eV, 2.33 eV, and 1.49 eV calculated using the HSE06 hybrid density functional, respectively). The calculated work functions and dielectric functions show that the four dynamically stable semiconductor structures are all high dielectric constant (high-k) materials, among which the β-P1̄ phase has the largest static dielectric constant (3.9 times that of SiO2). Furthermore, we explored band structures using the HSE06 functional and density of states (DOS) and the response of bandgaps to pressure using the PBE functional for the four new semiconductor configurations. The results show that the bandgap responses of the four structures exhibit significant differences when hydrostatic pressure is applied from 0 to 150 GPa.

Product family configuration optimisation considering after-sale service: an adaptive quantum evolutionary algorithm approach

ABSTRACT By providing after-sale services, manufacturing companies can gain more chances to interact with customers, improve brand preference and loyalty, occupy new bases in the value chain, and create more cross-selling opportunities. Considering after-sale service, we propose a new method for product family configuration optimisation in this study. Mathematical models for product family configuration optimisation are established with the objective of maximising the overall profit of the family of products and after-sale services. An adaptive quantum evolutionary algorithm (AQEA) is developed to solve the established optimisation models. An example of an e-book reader product is used to illustrate the proposed method in a case study. An industry case study shows that: (1) the proposed method outperforms the traditional method; (2) the proposed AQEA has better performance when compared with the other three meta-heuristic algorithms; and (3) the proposed method is more useful for industrial cases where the proportion of service profit is relatively large.

An Evolutionary Algorithmic Approach for Improving the Success Rate of Selective Assembly through a Novel EAUB Method

This work addresses an evolutionary algorithmic approach to reduce the surplus pieces in selective assembly to increase success rates. A novel equal area amidst unequal bin numbers (EAUB) method is proposed for classifying the parts of the ball bearing assembly by considering the various tolerance ranges of parts. The L16 orthogonal array is used for identifying the effectiveness of the proposed EAUB method through varying the number of bins of the parts of an assembly. Because of qualities such as minimal setting parameters, ease of understanding and implementation, and rapid convergence, the moth–flame optimization (MFO) algorithm is put forward in this work for identifying the optimal combination of bins of the parts of an assembly toward maximizing the percentage of the success rate of making assemblies. Computational results showed a 5.78% improvement in the success rate through the proposed approach compared with the past literature. The usage of the MFO algorithm is justified by comparing the computational results with the harmony search algorithm.

High-Pressure Phases of SnO and PbO: A Density Functional Theory Combined with an Evolutionary Algorithm Approach.

Tin monoxide, SnO, and its analog, lead monoxide, PbO, have the same tetragonal P4/nmm structure, shaped by nonbonding dispersion forces and lone pairs. The high-pressure phases of SnO and PbO have been explored in several experimental and theoretical studies, with conflicting results. In this study, the high-pressure structures of SnO and PbO are investigated using density functional theory calculations combined with an evolutionary algorithm to identify novel high-pressure phases. We propose that the monoclinic P21/m SnO and orthorhombic Pmmn PbO phases, which are metastable at 0 GPa, are a slight rearrangement of the tetragonal P4/nmm-layered structure. These orthorhombic (and their closely related monoclinic) phases become more favored than the tetragonal phase upon compression. In particular, the transition pressures to the orthorhombic γ-phase Pmn21 of SnO/PbO and the monoclinic phase P21/m of SnO are found to be consistent with experimental studies. Two new high-pressure SnO/PbO polymorphs are predicted: the orthorhombic Pbcm phase of SnO and the monoclinic C2/m of PbO. These phases are stabilized in our calculations when P > 65 GPa and P > 50 GPa, respectively. The weakening of the lone pair localization and elastic instability are the main drivers of pressure-induced phase transitions. Modulations of the SnO/PbO electronic structure due to structural transitions upon compression are also discussed.

A hybrid evolutionary algorithm approach for estimating the throughput of short reliable approximately balanced production lines

A hybrid evolutionary algorithm approach for estimating the throughput of short reliable approximately balanced production lines

Quantum inspired meta-heuristic approach for optimization of genetic algorithm

An optimization search technique evaluates the potential solution and frames the better solution by any of its befitting categories such as quantum-based approaches, meta-heuristic-based approaches, and multi-objective-based approaches. The Quantum Inspired Meta-Heuristic (QIMH) Algorithm is a synergistic combination of evolutionary algorithm, quantum computing, meta-heuristic, and multi-objective approaches to solve hard, exhaustive problems, and to get the best optimization. Techniques such as intensification or exploitation and diversification or exploration are useful in analyzing the optimum solution for the QIMH algorithm. The current research with QIMH algorithm provides a methodology for real time corona virus analysis from its prodigious parameters.

Multiple circle detection in images: a simple evolutionary algorithm approach and a new benchmark of images

Multiple circle detection in images: a simple evolutionary algorithm approach and a new benchmark of images

Promoting Electric Vehicle Cell Innovation Diffusion Considering Patent Licensing Strategy: A Combination of Evolutionary Game and Optimization Algorithm Approach

Electric vehicle cell industry is an emerging area with fierce competition on technical innovation, in which the patent holder can choose different innovation diffusion options to maximize the return; however, the strategy is unclear in certain scenarios. We tried to explain the question of how to maximize the patent holder’s return by appropriate patent license strategy to promote EV cell innovation diffusion, when competition and patent licensing relationship exist in the supply chain. A multistage and multichannel diffusion model of EV cell comprising the patent holder, EV cell producer and EV producers is developed; the evolutionary game is analyzed considering the competition among same stage players and patent licensing relationship among different stage players; and an optimization algorithm is introduced to find the maximum weighted object function of the patent holder. We established the multistage and multichannel diffusion model and found a nonlinear complex relationship between patent holder object function and the key factors including patent royalty pricing and innovation advantage coefficient; in addition, an optimization algorithm is developed based on adopters’ decision-making related with competition and patent licensing.

Analysing the role of roof mounted BIPV system optimization on decreasing the effect of duck curve in Perth, Western Australia: An experimental case study

This paper optimizes roof shape and associated Building Integrated Photovoltaic (BIPV) mounting strategies based on the newly announced feed-in tariff pricing range in Western Australia to confirm hypothesised benefits and potential influence on grid consumption reduction. Through analysis of a residential building in Perth, the study looks at: the current status of consumption, solar production, and energy balance. An Evolutionary Algorithm (EA) approach is deployed to optimize roof shape for each month and one whole year, with the aim of maximizing financial return by selling extra produced power to the grid. We have found that the optimized form has a tilt angle of 35.9° from horizontal and an azimuth angle of 47° toward the northwest. To compare the effectiveness of the newly announced tariff pricing, an analysis of economic benefits has been carried out comparing the ability of base case and optimized roof shapes to balance production and consumption surpluses and deficits. The outcome exemplifies the positive effect of an optimized roof shape that could supply 16% higher solar production than the base case shape. The current work provides practical guidelines for engineers, architects and researchers to design more suitably built environments.

Mixed Integer Programming versus Evolutionary Computation for Optimizing a Hard Real-World Staff Assignment Problem

Assigning staff to engagements according to hard constraints while optimizing several objectives is a task encountered by many companies on a regular basis. Simplified versions of such assignment problems are NP-hard. Despite this, a typical approach to solving them consists of formulating them as mixed integer programming (MIP) problems and using a stateof-the-art solver to get solutions that closely approximate the optimum.In this paper, we consider a complex real-world staff assignment problem encountered by the professional service company KPMG, with the goal of finding an algorithm that solves it faster and with a better solution than a commercial MIP solver. We follow the evolutionary algorithm (EA) metaheuristic and design a search heuristic which iteratively improves a solution using domain-specific mutation operators. Furthermore, we use a flow algorithm to optimally solve a subproblem, which tremendously reduces the search space for the EA.For our real-world instance of the assignment problem, given the same total time budget of 100 hours, a parallel EA approach finds a solution that is only 1.7% away from an upper bound for the (unknown) optimum within under five hours, while the MIP solver Gurobi still has a gap of 10.5%.

Modular system of simulation modeling of conventional and robotic mining technologies

The paper describes the function chart of a simulation modeling system for the conventional or robotic mining technologies with drilling-and-blasting and cutting methods. The function chart is implemented using the evolutionary algorithm and simulation approach. The authors present the main program modules with generate simulation models and optimize the technology and management scenarios for the specific geological and geotechnical conditions. The structure of the data base on mining machines, their capacities, robotic automation alternatives and compatibility is presented.

A global optima search field division method for evolutionary algorithms

This study aims to investigate the deployment of a proposed search field division method (SFDM) within evolutionary algorithms (EAs) to enhance the capability of searching for the global optima in nonlinear problems. The proposed technique is benchmarked against the following eight widely-used single-modal, multi-modal, and unimodal benchmark functions: Sphere, Rosenbrock, Rastringin, Griewank, Ackley, Fletcher, Quartic, and Schwefel functions, and the outcome is compared to their standard EAs counterparts to validate the effectiveness of the deployed approach in EAs. In the proposed method, we apply three low, medium, high field divisions (1, 2, and 5) dimensions on nine different EAs simultaneously with two different scenarios, 10 and 100 variables, to reach the optimal solution. Then for the validity of our proposed SFDM technique, we examined the exploration-exploitation search space rates and diversity behavior. The results of the implementation of SFDM on eight benchmark test functions show that the consideration of dimensions using SFDM for EAs improves the outcomes of all nine tested EAs. In our proposed method, we find better compatibility with the integration of SFDM in the Particle Swarm Optimization Algorithm concerning searching for the optimum solution relative to the other EAs. Highlights A novel Search Field Division Method (SFDM) for evolutionary algorithms. Applying three field divisions (1, 2, and 5) dimensions on nine different Evolutionary Algorithms (EAs) simultaneously to reach the optimal global solution. The implementation of SFDM on eight benchmark functions shows significant improvement in all nine tested EAs. In our proposed method, we find the Particle Swarm Optimization Algorithm has better compatibility with respect to the other EAs.