International Journal of Bio-Inspired Computation
These articles have been peer-reviewed and accepted for publication but are pending final changes, are not yet published and may not appear here in their final order of publication until they are assigned to issues. Therefore, the content conforms to our standards but the presentation (e.g. typesetting and proof-reading) is not necessarily up to the Inderscience standard. Additionally, titles, authors, abstracts and keywords may change before publication. Articles will not be published until the final proofs are validated by their authors.
Forthcoming articles must be purchased for the purposes of research, teaching and private study only. These articles can be cited using the expression "in press". For example: Smith, J. (in press). Article Title. Journal Title.
Articles marked with this shopping trolley icon are available for purchase - click on the icon to send an email request to purchase.
Register for our alerting service, which notifies you by email when new issues are published online.
International Journal of Bio-Inspired Computation (26 papers in press)
Special Issue on: Artificial Intelligence Facilities Smart Cities Development
Abstract: Real-time people counting based on videos is one of the most popular projects in the construction of smart cities. To develop an accurate people counting approach, deep learning can be used as it greatly improves the accuracy of machine learning based approaches. To this end, we have previously proposed an accurate YOLO (You Only Look Once) based People Counting approach, dubbed YOLO-PC. However, the model of YOLO-PC was very large with an excessive number of parameters, thus it requires large storage space on the device and makes transmission on Internet a time consuming task. In this paper, a new real-time people counting method named as Squeeze YOLO-based People Counting (S-YOLO-PC) is proposed. S-YOLO-PC uses the fire layer of SqueezeeNet to optimize the network structure, which reduces the number of parameters used in the model without decreasing its accuracy. Based on the obtained the experimental results, S-YOLO-PC reduces the number of model parameters by 11.5% and 9% compared to YOLO and YOLO-PC, respectively. S-YOLO-PC can also detect and count people with 41 frames per second (FPS) with the average precision (AP) of person of 72%.
Keywords: Model compression; People counting; Boundary-selection; YOLO; SqueezeNet.
SmartGC: A Software Architecture for Garbage Collection in Smart Cities
by Miguel Ramalho, Rosaldo Rossetti, Nelio Cacho, Arthur Souza
Abstract: With populations in cities increasing in a very accelerated pace, the problem of collecting and handling the waste produced becomes a major concern to governmental authorities. Indeed, the amount of garbage they create is increasing even faster than their populations, worsening the problem and turning garbage collection into a very challenging task. In this paper, we see garbage collection through the spectacles of the emerging concept of Smart Cities, accounting for new performance measures defined on the grounds of sustainability, energy efficiency, optimum resource allocation, and low carbon emission and footprint. We thus devise a smart garbage collection management system, coined SmartGC, whose architecture is detailed and explained. Abstracting out garbage collection from a smart mobility perspective, the underlying methodology supporting the proposed architecture relies on the concept of Artificial Transportation Systems. For the sake of demonstration, we have implemented a routing strategy to generate improved itineraries accounting for the content of garbage containers, which are continuously monitored through IoT-based smart meters. Also, we discuss on how the architecture is instantiated and integrated into the smart city agenda of Natal, a medium-size capital in Northeastern Brazil.
Keywords: smart cities; garbage collection; artificial transportation systems; software architecture.
Time-to-Contact Control: Improving Safety and Reliability of Autonomous Vehicles
by Liang Wang, Berthold K.P. Horn
Abstract: Under traditional car-following control, i.e. human drivers' behavior, the stability condition of traffic system is not satisfied in general. For safety and reliability of autonomous vehicles, additional danger warning system must be used in the adaptive cruise control system to prevent inevitable potential collisions. One reasonable quantity of evaluating potential collisions is time to contact (TTC): how soon will potential collision occur? In this paper, we provide TTC feedback control to improve safety and reliability of autonomous vehicles, and show the effectiveness of TTC feedback. TTC can be estimated by machine vision technics with single uncelebrated camera (i.e. passive sensors). We provide detailed mathematical analysis and algorithmic implementation. The machine vision based TTC algorithm is pretty fast such that the whole system can be implemented on Android smart phones running in real-time. Moreover, it's not trial to estimate relative velocity by differentiating the measured distance between cars with respect to time, because inevitable measurement noise in the distance measurements will be amplified by the derivative operation. The time-to-contact based algorithm provides an alternative approach to estimating the relative velocity, which can also be fused with measurements from other active sensors, if desired.
Keywords: Autonomous vehicles; time to contact (TTC); danger detection; machine visionp; android smartphones.
by Jiao-Hong Yi
Abstract: As a complicated high-dimensional optimization problem, path planning navigation problem for Uninhabited Combat Air Vehicles (UCAV) is to obtain a shortest safe flight route with different types of constrains under complicated combating environments. Monarch butterfly optimization (MBO) is a highly promising swarm intelligence algorithm. Since then, though it has successfully solved several challenging problems, MBO may be trapped into local optima sometimes. In order to improve the performance of MBO, quantum computation is firstly incorporated into the basic MBO algorithm, and a new quantum inspired MBO algorithm is then proposed, called QMBO. In QMBO, certain number of the worst butterflies are updated by quantum operators. In this paper, the UCAV path planning navigation problem is modeled into an optimization problem, and then its optimal path can be obtained by the proposed QMBO algorithm. In addition, B-Spline curves are utilized to further smoothen the obtained path and make it more feasible for UCAV. The UCAV path obtained by QMBO is compared with the basic MBO, and the experimental results show that QMBO can find much shorter path than MBO.
Keywords: Unmanned combat air vehicle; Path planning navigation; Monarch butterfly optimization; Quantum computation; B-Spline curve.
Dynamic Economic Emission Dispatch Problem with Renewable Integration Focusing on Deficit Scenario in India
by Hithu Anand, Rengaraj Ramasubbu
Abstract: Present scenario power grid has increased penetration of renewable energy sources (RESs). RESs are clean sources of energy and power production from them with minimal emission is a national target. Already, increased carbon footprint has put nations into jeopardy. Nowadays to study the benefits due to RESs in power system is of greater importance. Stochastic nature of RESs made it difficult to manage power dispatch scenario. Dynamic power demand added even more difficulty in obtaining real time economic schedule of generation dispatch. A test system of ten generator and emission dispatch with wind turbine (WT) and photo voltaic (PV) having dynamic load for 24 hours is economized. Stochastic method of particle swarm optimization (PSO) is compared with anti-predatory particle swarm optimization (APSO). It is identified that APSO method gives a better economy with reduced emission for the given problem.
Keywords: renewable energy; dynamic economic emission dispatch; wind power; solar power; anti-predatory particle swarm optimization.
Integrated deteriorating maintenance and patient scheduling for single medical device with heuristic algorithm
by Liu Qinming
Abstract: This paper aims to propose a two-phase model integrated patient scheduling and medical device maintenance to improve their reliability, reduce operating costs, and increase operating efficiency. In this paper, one patient scheduling problem with time-window deteriorating maintenance is studied. The objective is to minimize the maximum tardiness of all patients. First, a two-phase mathematical model is developed to characterize the problem. One model is used to solve the lower bound of the number of maintenance activities, and the other is used to obtain the patient scheduling solution. Then, one heuristic is developed for the problem. Finally, numerical experiments can be performed to indicate the efficiency and effectiveness of the proposed methods. The results show that the proposed methods have a better performance for the patient scheduling problem and can be able to obtain one good solution in a short computation time. Few studies have been carried out to integrate decisions between patient scheduling and device maintenance. Their considerations are either incomplete or not realistic enough. A more comprehensive and realistic two-phase model is proposed in this paper.
Keywords: Patient scheduling; Maximum tardiness; Medical device; Time-window deteriorating maintenance; Virtual maintenance.
Performance Analysis of Intrinsic Embedded Evolvable Hardware using Memetic and Genetic Algorithms
by Ranjith Chandrasekharan, Joy Vasantha Rani S.P
Abstract: This paper discusses the performance analysis of memetic and genetic algorithms (GA and MA) as the optimising strategy for the design of embedded evolvable hardware. The optimisation algorithm with the fitness evaluation searches for the best configuration to evolve the hardware model. Here, an experimental setup is carried to intrinsically evolve combinational circuits to test the performance of MA and GA. The complete evaluation and evolution is built on a single Virtex 6 (XC6VLX240T-1FFG1156) ML605 Evaluation Kit FPGA. A Virtual Reconfigurable Architecture (VRA) with the hardware fitness circuit is modelled as a second reconfigurable layer over the Field Programmable Gate Array (FPGA) to configure the target combinational logic. A FPGA soft core processor evaluates the search algorithm and the best solutions are utilised for the hardware evolution. The experimentation results showed that convergence and evolution time of MA was faster compared to GA when the search space was large. Thus, proving MA is a better option for large search space evaluations for evolvable hardware architectures.
Keywords: evolvable hardware; EHW; embedded evolvable hardware; evolutionary algorithm; genetic algorithm; memetic algorithm; MicroBlaze processor; VRA; Virtual Reconfigurable Architecture; evolution speed; evaluation time; combinational circuits; intrinsic evolution; Bio-Inspired algorithm.
A New Replica Placement Strategy Based on Multi-objective Optimization for HDFS
by Wang Yang, Zhang Qingfu, Li Yangyang, Dhish Saxena
Abstract: Distributed storage systems like the Hadoop Distributed File System (HDFS) constitute the core infrastructure of cloud platforms which are well poised to deal with big-data. An optimized HDFS is critical for effective data management in terms of reduced file service time and access latency, improved file availability and system load balancing. Recognizing that the file-replication strategy is key to an optimized HDFS, this paper focuses on the file-replica placement strategy while simultaneously considering storage and network load. Firstly, the conflicting relationship between storage and network load is analyzed and a bi-objective optimization model is built, following which a multi-objective optimization memetic algorithm based on decomposition (MOMAD) and its improved version are used. Compared to the default strategy in HDFS, the file-replica placement strategies based on multi-objective optimization provide more diverse solutions. And competitive performance could be obtained by the proposed algorithm.
Keywords: Hadoop; HDFS; replica placement; multi-objective optimization; memetic algorithm.
Adaptive Neighborhood Size Adjustment in MOEA/D-DRA
by Meng Xu
Abstract: The multiobjective optimization algorithms based on decomposition(MOEA/D) is a well-known multiobjective optimization algorithms(MOEAs). MOEA/D was proposed by Zhang and Li in 2007s. MOEA/D decomposes a multiobjective problem into a set of scalar single objective subproblems using the aggregation function and the evolutionary operator. The variant of the dynamic resource allocation strategy in MOEA/D(MOEA/D-DRA) has the outstanding performance on CEC2009, the MOEA/D-DRA using the strategy of resource allocation. It cares about the convergence and ignores the diversity. MOEA/D-DRA is very sensitive to the neighbourhood size. In this paper, we present a new enhanced MOEA/D-DRA strategy based on the adaptive neighbourhood size adjustment(MOEA/D-DRA) to increase the diversity. It focuses on the solutions density around of subproblems. The experiment results demonstrate that MOEA/D-ANA strategy performs the best compared with other five classical MOEAs on the CEC2009 test instances.
Keywords: MOEA/D; diversity; Neighborhood; CEC2009 test instances;.
Parallel Implementation of Genetic Algorithm on FPGA using Vivado High Level Synthesis
by Eman Alqudah, Amin Jarrah
Abstract: Genetic Algorithm (GA) is one of most popular evolutionary search algorithms that simulates natural selection of genetic evolution for searching solution to arbitrary engineering problems. However, it is computationally intensive and will become a limiting factor for evolving solution to most of the real life problems as it involves large number of parameters that needs to be determined. Fortunately, there are some parallel platforms such as Field-Programmable Gate Array (FPGA) that can be adopted to overcome this constrains by improving its latency. So, efficient parallel implementation of GA was proposed where each step of GA was exploited to improve its computational task. Moreover, many optimization and parallelization techniques were adopted and applied to achieve high speed up. The results show that 43 speed up is achieved compared with the typical one. Moreover, higher speed up can be achieved with larger input size.
Keywords: Genetic Algorithm (GA); Field Programmable Gate Array (FPGA); Vivado HLS tool; parallel architecture; optimization techniques.
by Redouane Boudjemaa, Diego Oliva, Fatima Ouaar
Abstract: A well-known metaheuristic is the Bat Algorithm (BA), which consists of an iterative learning process inspired by bats echolocation behaviour in searching for prays. Basically, the BA uses a predefined number of bats that collectively move on the search space to find the global optimum. This article proposes the Fractional L
Keywords: Fractional calculus; Bat algorithm; Lévy Flight; Nonparametric statistical tests.
A Modified Bat Algorithm With Torus Walk for Solving Global Optimization Problems
by Waqas Bangyal, Jamil Ahmad, Hafiz Tayyab Rauf
Abstract: Bat algorithm (BA) has been widely used to solve the diverse kind ofrnoptimization problems. In accordance with the optimization problem, a balance between the two major components exploitation and exploration plays a significant role in metaheuristic algorithms. These algorithms can influence diversity, convergence and help to find an optimal solution efficiently. Particularly, recognizing the importance of diversityrnand convergence, several researchers have worked on the performance for the improvement of meta-heuristic algorithms. BA faces one of the major issues in high dimensions for numerical optimization problems. In our work, we proposed a new variant of BA by introducing the Torus walk (TW-BA) to solve this issue. To improve the local search capability instead of using the standard uniform walk, Torus walk is incorporated in this paper. Additionally, chaotic inertia weight is introduced to manage the local search capability. The simulation results performed on nineteen standard benchmark functions depicts the efficiency and effectiveness of TW-BA compared with the traditional BA, directional Bat Algorithm (dBA), Particle Swarm Optimization (PSO), Cuckoo Search via L`evy flights (CS), Harmony Search Algorithm(HS), Classical Differential Evolution (DE) and Standard Genetic Algorithm (GA). The experimental results show that the proposedrntechnique performed exceptionally better than the standard technique. Moreover, the outcome of our work presents a foresight on how the proposed technique highly impacts on the value of cost function, convergence, and diversity. The promising experimental result suggests the superiority of the proposed technique.
Keywords: Torus Walk; Chaotic Inertia Weight; Exploitation; Exploration.
An enhanced genetic algorithm for the distributed assembly permutation flowshop scheduling problem
by Xin Zhang, Xiangtao Li, Minghao Yin
Abstract: The distributed assembly permutation flowshop scheduling problem (DAPFSP) is a new generalization of the distributed permutation flowshop scheduling problem (DPFSP) and the assembly flowshop scheduling problem (AFSP), aiming to minimise makespan. This production mode is more complicated and competitive in the real production process and includes two phases: production and assembly. Firstly, the production is conducted in several identical factories, and the production in each factory can be considered to a permutation flowshop scheduling problem (PFSP) with multi-machines. Then, the jobs produced in the first stage are assembled into final products. An enhanced population-based meta-heuristic genetic algorithm (GA) is proposed for this problem. A greedy mating pool is designed to select promising parents in the selection operation, and an effective crossover strategy is designed based on the local search for speeding up convergence. To enhance the exploitation capability, several different local search strategies are incorporated into the algorithm, which are based on two neighborhood structures. The exhaustive experiment and statistical analysis show that the proposed algorithms outperform the existing algorithms.
Keywords: Distributed assembly scheduling; Permutation flowshop; Meta-heuristic; Genetic algorithm; Crossover; Local search.
A hybrid bio-inspired optimization approach for wirelength minimization of hardware tasks placement in Field Programmable Gate Array devices
by Premalatha Balasubramaniam, Uma Maheswari S
Abstract: In Computer Aided Design (CAD) flow of VLSI Circuits, Placement Process is an NP-Complete problem which requires an optimization approach to obtain the system performance better. The objective of the placement process is to reduce the wire length between the placed tasks with zero overlap. Fast response and better convergence algorithms are required to meet these desires. In this regard, bio-inspired optimization algorithms such as Genetic algorithm (GA) and Particle Swarm Optimization (PSO) algorithm have been considered in this paper to solve the issue. PSO is a robotic bio-inspired optimization procedure based on the swarm intelligence. The genetic algorithm is also a bio-inspired process based on natural selection and generates the solution to optimization problems. The fitness function is considered as cost function which is formulated to minimize the wire length with zero overlap. By using the salient features of these two algorithms, the optimized solution for placement problem have been obtained. In the first phase, the concept of GA has been applied to obtain optimized wire length with zero overlap between the task and the solution from Genetic Algorithm is taken as an input to the Particle Swarm Optimization to attain the enhanced optimized result. For experimentation, various Directed Data Flow Graphs (DDFGs) are randomly generated and the comparison is made between the individual GA, PSO and hybrid (GA-PSO) method. The hybrid approach using GA-PSO produces better experimental results in wire length minimization and hence outperforms than the others.
Keywords: Hardware tasks placement; NP-Complete Problem; wirelength minimization; Reconfigurable FPGAs; bio-inspired optimization approach; Genetic Algorithm; GA; Particle Swarm Optimization; PSO.
Bat Algorithm with Weibull Walk for Solving Global Optimization and Classification Problems
by Hafiz Tayyab Rauf, Muhammad Hadi, Abdur Rehman
Abstract: Bat algorithm (BA) becomes the most widely employed meta-heuristic algorithm to interpret the diverse kind of optimization and real-world classification problems. Toward accordance between the classification and optimization problem, the stability and the endurance among the two principal components exploitation and exploration impersonate a meaningful role in the family of meta-heuristic algorithms. BA suffers from one of the influential challenges called local minima for numerical optimization and real-world classification problems. In this study, we carried out two modifications in the original BA and proposed a modified variant of BA called Bat Algorithm with Weibull walk (WW-BA) to solve the premature convergence issue. The first modification involves the introduction of Weibull descending inertia weight for updating the velocity of bats. The second modification approach updates the local search strategy of BA by replacing the Random walk with the Weibull Walk. The simulation performed on nineteen standard benchmark functions represents the competence and effectiveness of WW-BA compared with the traditional BA, Cuckoo Search via L
Keywords: Bat Algorithm; Premature Convergence; Exploration; Exploitation; Weibull Walk; Inertia weight.
A privacy-preserving recommendation method based on multi-objective optimization for mobile users
by Chonghuan Xu
Abstract: Recommender systems have proven to be an effective technique to deal with information overload and mislead problems by helping users get useful and valuable information or objects from massive data. However, exploiting users preferences with recommendation algorithms lead to serious privacy risks, especially when recommender service providers are unreliable. An ideal recommender system should be both accurate, diverse and security. In this paper, we propose a private recommendation method which consists of a private collaborative filtering algorithm and a multi-objective evolutionary algorithm for mobile users. Experimental results demonstrate that even though the mobile users preferences are significantly obfuscated, our method is effective in terms of recommendation accuracy and diversity.
Keywords: Recommender systems;Multi-objective optimization;Differential privacy;Mobile users.
Data-driven Pollution Source Location Algorithm in Water Quality Monitoring Sensor Networks
by Xuesong Yan
Abstract: Water pollution prevention has been a widely concerned issue for the safety of human lives. To this end, water quality monitoring sensors are introduced in the water distribution systems. Due to the limited budget, it is impossible to deploy sensors everywhere but a small number of sensors are deployed. From the sparse sensor data, it is important, but also challenging, to find out the pollution source location. Traditional methods may suffer from local optimum trapping or low localization accuracy. To address such problems, we propose a cooperative intelligent optimisation algorithm based pollution source location algorithm, which is a data-driven approach in simulation-optimisation paradigm. Through open-source EPANET simulator based experiments, we find out our proposed data-driven algorithm can effectively and efficiently localize the pollution location, as well as the pollution injection starting time, duration and mass.
Keywords: sensor networks; pollution source location; simulation optimisation; cooperative optimisation algorithm.
Recognition of driver emergency braking behavior based on support vector machine optimized by memetic algorithm
by Shenpei Zhou, Bingchen Qiao, Haoran Li, Bin Ran
Abstract: Surface electromyography (sEMG) is one of the main information sources of human motion detection and has been widely used. The lower limb sEMG signal is introduced into the recognition model of driver emergency braking behavior, and the features from time domain, frequency domain and model parameters are extracted to construct a feature vector. In addition, to improve recognition accuracy, the data from conventional braking and accelerated shifting behaviors similar to the characteristics of emergency braking are synchronously collected. These three driving behaviors are identified by using support vector machine (SVM), and a memetic algorithm (MA) based on particle swarm optimization and hill climbing algorithm is proposed to optimize the parameters of SVM. The results show that the model based on SVM optimized by MA has better classification performance than that without optimization. The final recognition rate of emergency braking behavior of same individual is up to 92.3%, and that of different individuals can reach 85.6%. Moreover, the system can detect emergency braking 220 ms earlier than operating brake pedal. At 100 km/h driving speed, this amounts to reducing the braking distance by 6.1 m.
Keywords: sEMG; emergency braking; SVM; parameter optimization; memetic algorithm; particle swarm optimization; hill climbing algorithm.
Performance-Aware Deployment of Streaming Applications in Distributed Stream Computing Systems
by Dawei Sun, Shang Gao, Xunyun Liu, Fengyun Li, Rajkumar Buyya
Abstract: Performance-aware deployment of streaming applications is one of the key challenging problems in distributed stream computing systems. We proposed a performance-aware deployment framework (Pa-Stream) for distributed stream computing systems. By addressing the important aspects of the framework, this paper makes the following contributions: (1) Investigated the performance-aware deployment of a streaming application over distributed and heterogeneous computing nodes, and provided a general application deployment model. (2) Demonstrated a streaming applications deployment scheme by proposing an artificial bee colony strategy that deploys applications vertices onto the best set of computing nodes; an incremental online redeployment strategy was used to redeploy the running application. (3) Developed and integrated Pa-Stream into Apache Storm platform. (4) Evaluated the fulfillment of low latency and high throughput objectives in a distributed stream computing environment. Experimental results demonstrate that the proposed Pa-Stream provided effective performance improvements on latency, throughput and resource utilization.
Keywords: performance awareness; application deployment; stream computing; artificial bee colony algorithm; distributed system.
Improved Density Peaks clustering based on Firefly Algorithm
by Jia Zhao, Jingjing Tang, Aiye Shi, Tanghuai Fan, Lizhong Xu
Abstract: The cutoff distance of the Density Peaks Clustering (DPC) algorithm need to be set manually; the two local densities defined by the algorithm have a large difference in the clustering effect on the same dataset; and the calculation of the decision value does not take into account the different effects of the sample local density and relative distance on the cluster center, so the applicability and clustering effect of the DPC algorithm is not good. To address the issue, the paper proposes a new method, called improved density peaks clustering based on firefly algorithm. It combines the cutoff kernel and the Gaussian kernel defined by the DPC algorithm, and balances the effects of the two kernels by the weighting factor, thus a new definition of local density is proposed. Meanwhile, considering the different influences of local density and relative distance on decision values, a cluster-like center evaluation criterion based on local density and relative distance of preference coefficient is constructed. In order to determine the parameters of the cutoff distance, weighting factor and preference coefficient, the three parameters are optimized by the firefly algorithm with the Rand Index as the objective function. The experiment results show that the performance of the proposed method on synthetic datasets and real datasets is better than DPC and its variants, and realizes the parameter adaptation to the dataset with arbitrary shapes and sizes.
Keywords: Density peaks clustering; Firefly algorithm; cutoff distance; Weighting factor; Preference coefficient.
NSGA-III algorithm with maximum ranking strategy for many-objective optimization
by Fei Xue, Di Wu
Abstract: In recent years, a non-dominated sorting genetic algorithm III (NSGA-III) based on decomposition strategy had been extensively studied. However, there are still problems of lower pareto selection pressure and insufficient diversity maintenance mechanism. To address these problems, NSGAIII algorithm with maximum ranking strategy (NSGAIII-MR) is proposed. In this algorithm, the convergence and diversity distance are balanced by adaptive parameter settings to achieve better performance. The maximum ranking strategy exploits the perpendicular distance from the solution to the weight vector to increase pareto selection pressure. Moreover, the diversity of population is maintained with the reference point strategy to guide the solutions closer to the real pareto front. Comparing with NSGAIII, the NSGAIII-MR algorithm enhances selection pressure and has good convergence and diversity performance. Also, the performance of algorithm is verified by comparing with other state-of-the-art evolutionary algorithms on the benchmark problems and the NSGAIII-MR is competitive.
Keywords: convergence; maximum ranking strategy; diversity; many-objective evolution algorithm.
Variable-Grouping-based Exponential Crossover for Differential Evolution Algorithm
by Shu Yang, Qiuling Huang, Laizhong Cui, Kunkun Xu, Zhong Ming, Zhenkun Wen
Abstract: The performance of differential evolution (DE) algorithm largely depends on its crossover operator, whose substantive characteristics is to make the algorithm search in a subspace of the original search space. Different crossover operators use different subspace divisions, and how to choose a suitable crossover operator for a specific optimization problem is still an open issue. This paper proposes variable-grouping-based exponential crossover (VGExp), where all variables are divided into multiple groups based on interaction information, and the variables that are mutated simultaneously have a high probability of coming from the same group. Moreover, the solutions can improve the accuracy of the variable grouping and provide initial guidance for optimization. Therefore, the proposed VGExp seamlessly combines variables grouping technique and differential evolution. The experiment results based on 30 CEC2014 test problems show that VGExp can improve the performance of most DE variants, and it is also better than other well-developed crossover operators.
Keywords: Differential evolution; exponential crossover; variable grouping; variable interaction.
Archived Elitism in Evolutionary Computation: Towards Improving Solution Quality and Population Diversity
by Maxim Dulebenets
Abstract: Many Evolutionary Algorithms, developed for solving complex optimization problems, deploy the elitist strategy. The elitist strategy ensures that a group of the fittest individuals will be transferred to the next generation before performing any algorithmic operations. In general, elitism allows improving the algorithmic performance in terms of solution quality. However, transferring a group of the fittest individuals to the next generation will increase the selection pressure and significantly limit chances of the newly created offspring chromosomes to survive. In order to address the latter drawbacks, this study proposes and evaluates a number of alternative archive-based elitist strategies, where the fittest individuals are stored in the archive and transferred from that archive into the population based on certain rules. The computational experiments are conducted for the unrelated machine scheduling problem, where the total job processing cost is minimized. The results indicate that the proposed strong archived elitism strategy, which samples the best individual discovered from the archive in every generation, outperforms the other elitist strategies in terms of the objective function values by up to 8.29% over the considered problem instances. Moreover, the strong archived elitism strategy improves the population diversity, which further facilitates the explorative capabilities of the algorithm.
Keywords: Evolutionary computation; optimization; elitist strategies; machine scheduling problems; population diversity; strong archived elitism; solution quality.
An Enhanced Breeding Swarms Algorithm for High Dimensional Optimizations
by Jon A. Hansen, Jørgen Sund, Dylan Tollemache, Ali Arefi, Ghavameddin Nourbakhsh
Abstract: This paper proposes a metaheuristic optimisation algorithm named Enhanced Breeding Swarms (EBS), which combines the strengths of Particle Swarm Optimization (PSO) with those of Genetic Algorithm (GA). In addition, EBS introduces three modifications to the original Breeding Swarms to improve the performance and the accuracy of the optimisation algorithm. These modifications are applied on the acceptance criteria based on the Improved Glowworm Swarm Optimization, Velocity Impact Factor, and the mutation operator. The EBS algorithm is tested and compared against GA, PSO, and original BS algorithms, using unrotated and rotated six recognised optimisation benchmark functions. Results indicate that the EBS outperforms GA, PSO, and BS in most cases in terms of accuracy and speed of convergence, especially when the dimension of optimisation increases. As an application of the proposed EBS algorithm, a load flow analysis on a 6-bus network is performed, and the comparison results against another heuristic algorithm and the Newton-Raphson are reported.
Keywords: Enhanced Breeding Swarms; Particle Swarm Optimization; Generic Algorithm; Metaheuristic Optimization; Improved Glowworm Swarm Optimization; Computational Intelligence.
Genetic Optimized Serial Hierarchical Fuzzy Classifier for Breast Cancer Diagnosis
by XIAO ZHANG, Enrique Onieva, Asier Perallos, Eneko Osaba
Abstract: Accurate early-stage detection and medical diagnosis of breast cancer can improve the survival rates, in which fuzzy rule-base system (FRBS) has been a promising classification system to detect breast cancer. However, the existing classification systems involves large number of input variables for training and produces a large number of fuzzy rules, which lead to high system complexity and barely acceptable classification accuracy. In this paper, we present a new classification system by means of fuzzy logic and genetic algorithm to classify the breast cancer disease as benign or malignant. More specifically, a serial hierarchical FRBS is optimized by genetic algorithm, which incorporate lateral tuning of the membership functions and optimization of the rule base to improve the classification accuracy. The genetic optimized serial hierarchical structure of FRBS allows to select and rank the input variables, which reduce the system complexity and distinguish the importance of the attributes in datasets. To evaluate the proposed system, we conduct an experimental study on Original Wisconsin Breast Cancer Database and Wisconsin Breast Cancer Diagnostic Database from UCI Machine Learning Repository. It shows that the proposed system can achieve a promising classification performance for breast cancer diagnosis while reducing the size of involved input variables and rule base.
Keywords: Genetic Algorithm; Fuzzy Logic; Classification System; Breast Cancer Diagnosis; Variable Selection.
A Bee Colony Optimization Algorithm with a Sequential-Pattern-Mining-based Pruning Strategy for the Traveling Salesman Problem
by Shin Siang Choong, Li-Pei Wong, Malcolm Yoke Hean Low, Chin Soon Chong
Abstract: Bees perform waggle dance in order to communicate the information of food source to their hive mates. This unique foraging behaviour has been computationally realized as an algorithmic tool named the Bee Colony Optimization (BCO) algorithm to solve different types of Combinatorial Optimization Problems such as Traveling Salesman Problem (TSP). In order to enhance the performance of BCO, local optimization can be integrated. However, local optimization incurs high processing overhead especially when all solutions are allowed to undergo the local optimization. To reduce the high processing overhead, two existing pruning strategies, i.e. Frequency-based Pruning Strategy (FBPS) (Wong et al., 2009b) and Frequent-closed-pattern-based Pruning Strategy (FCPBPS) (Wong & Choong, 2015) were proposed to prohibit a subset of solutions from undergoing the local optimization. The rationale of these pruning strategies is to allow only the solutions which contain a significant amount of frequent building blocks to perform local optimization. This paper proposes a new pruning strategy based on the top-k sequential patterns mining (TKS) algorithm. Specifically, TKS is employed to identify the frequent building blocks along the optimization process. A total of 19 selected symmetric TSP benchmark problem instances ranging from 318 cities to 1291 cities were used as the test bed of this study. Based on the experimental results, the proposed pruning strategy shows a significant reduction in terms of the computational time to yield TSP solutions with similar tour length as compared with two state-of-the-art approaches.
Keywords: meta-heuristic; local search; data mining; sequential pattern mining; frequency-based pruning strategy; frequent-close-pattern-based pruning strategy; combinatorial optimization.