Mostafa Z. Ali

Multi-population differential evolution with balanced ensemble of mutation strategies for large-scale global optimization

Differential evolution (DE) is a simple, yet very effective, population-based search technique. However, it is challenging to maintain a balance between exploration and exploitation behaviors of the DE algorithm. In this paper, we boost the population diversity while preserving simplicity by introducing a multi-population DE to solve large-scale global optimization problems. In the proposed algorithm, called mDE-bES, the population is divided into independent subgroups, each with different mutation and update strategies. A novel mutation strategy that uses information from either the best individual or a randomly selected one is used to produce quality solutions to balance exploration and exploitation. Selection of individuals for some of the tested mutation strategies utilizes fitness-based ranks of these individuals. Function evaluations are divided into epochs. At the end of each epoch, individuals between the subgroups are exchanged to facilitate information exchange at a slow pace. The performance of the algorithm is evaluated on a set of 19 large-scale continuous optimization problems. A comparative study is carried out with other state-of-the-art optimization techniques. The results show that mDE-bES has a competitive performance and scalability behavior compared to the contestant algorithms.

An ensemble sinusoidal parameter adaptation incorporated with L-SHADE for solving CEC2014 benchmark problems

An effective and efficient self-adaptation framework is proposed to improve the performance of the L-SHADE algorithm by providing successful alternative adaptation for the selection of control parameters. The proposed algorithm, namely LSHADE-EpSin, uses a new ensemble sinusoidal approach to automatically adapt the values of the scaling factor of the Differential Evolution algorithm. This ensemble approach consists of a mixture of two sinusoidal formulas: A non-Adaptive Sinusoidal Decreasing Adjustment and an adaptive History-based Sinusoidal Increasing Adjustment. The objective of this sinusoidal ensemble approach is to find an effective balance between the exploitation of the already found best solutions, and the exploration of non-visited regions. A local search method based on Gaussian Walks is used at later generations to increase the exploitation ability of LSHADE-EpSin. The proposed algorithm is tested on the IEEE CEC2014 problems used in the Special Session and Competitions on Real-Parameter Single Objective Optimization of the IEEE CEC2016. The results statistically affirm the efficiency and robustness of the proposed approach to obtain better results compared to L-SHADE algorithm and other state-of-the-art algorithms.

Mining the Social Fabric of Archaic Urban Centers with Cultural Algorithms

Applying a suite of tools from artificial intelligence and data mining to existing archaeological data from Monte Alban, a prehistoric urban center, offers the potential for building agent-based models of emergent ancient urban centers. The authors use decision trees to characterize location decisions made by early inhabitants at Monte Alban, a prehistoric urban center, and inject these rules into a socially motivated learning system based on cultural algorithms. They can then infer an emerging social fabric whose networks provide support for certain theories about urban site formation. Specifically, we examine the period of occupation associated with the emergence of this early site. Our goal is to generate a set of decision rules using data-mining techniques and then use the cultural algorithm toolkit (CAT) to express the underlying social interaction between the initial inhabitants.

A Differential Evolution algorithm with success-based parameter adaptation for CEC2015 learning-based optimization

Developing efficient evolutionary algorithms for solving learning-based real-parameter single objective optimization is a very challenging and essential task in many real applications. This task involves finding the best optimal solution with least computational cost, avoiding premature convergence. This paper proposes a new efficient Differential Evolution algorithm with success-based parameter adaptation with resizing population space. We introduce a new technique to adapt the control parameters which uses a memory-based structure of previous successful settings. Moreover, the population size is adapted linearly to find the most suitable size which helps to guide the search in each optimization loop. The proposed algorithm is tested on the benchmarks of the CEC2015 real parameter single objective competition. The results affirm the efficiency and robustness of our approach to reach good results.

A novel class of niche hybrid Cultural Algorithms for continuous engineering optimization

This paper proposes a novel class of niche hybrid Cultural Algorithms for solving engineering problems with continuous design variables. The first algorithm, Niche Cultural Algorithm (NCA), embeds niching within the cultural framework to maintain multiple groups within the population of agents in order to locate multiple optima. The second algorithm, hybridizes niche Cultural Algorithms with Tabu search (H-NCA). This technique offers a novel architecture of hybrid approaches, which combines Niche Cultural Algorithms (NCA) with Tabu search (TS). The proposed hybridization scheme enables the algorithm to overleap local optima and improve performance. The third algorithm, Improved Hybrid Niche Cultural Algorithms (IH-NCA), is employed to enhance convergence rate and accuracy of H-NCA with fewer computations. In IH-NCA, the algorithm switches between two selection strategies based on roulette wheel and stochastic tournament selection. This enhances the algorithms ability to further escape stagnation and premature convergence with varying stochastic noise and selection pressure. Simulations were performed over miscellaneous engineering optimization problems that include minimization of constrained functions and structural engineering optimization. A comparative study is carried out with other state-of-the-art optimization techniques. The findings affirm the efficiency and robustness of the new methodologies over the other existing relevant approaches.

Embedding a social fabric component into cultural algorithms toolkit for an enhanced knowledge‐driven engineering optimization

Purpose – The purpose of this paper is to introduce the notion of a social fabric (SF) in which the expression of knowledge sources (KS) in cultural algorithms (CA) can be distributed through the population. The SF influence function is applied to the solution of selected complex engineering problems and it is shown that different parameter combinations for the SF influence function can affect the rate of solution. This enhanced approach is compared with previous approaches.Design/methodology/approach – KS are allowed to influence individuals through a network. From a theoretical perspective, individuals in the real world are viewed as participating in a variety of different networks. Several layers of such networks can be supported within a population. The interplay of these various network computations is designated as the “social fabric.” Using this new influence function, when an individual is to be modified, one KS is selected to perform the modification at each generation. The selection process is d...

CADE: A hybridization of Cultural Algorithm and Differential Evolution for numerical optimization

Abstract Many real-world problems can be formulated as optimization problems. Such problems pose a challenge for researchers in the design of efficient algorithms capable of finding the best solution with the least computational cost. In this paper, a new evolutionary algorithm is proposed that combines the explorative and exploitative capabilities of two evolutionary algorithms, Cultural Algorithm (CA) and Differential Evolution (DE) algorithm. This hybridization follows the HTH (High-level Teamwork Hybrid) nomenclature in which two meta-heuristics are executed in parallel. The new algorithm named as CADE, manages an overall population which is shared between CA and DE simultaneously. Four modified knowledge sources have been used in proposed CA which are: topographical, situational, normative and domain. The role of the used acceptance function in belief space is to select the knowledge of the best individuals to update the current knowledge. A novel quality function is used to determine the participation ratio for both CA and DE, and then a competitive selection takes place in order to select the proportion of function evaluations allocated for each technique. This collaborative synergy emerges between the DE and CA techniques and is shown to improve the quality of solutions, beyond what each of these two algorithms alone. The performance of the algorithm is evaluated on a set of 50 scalable optimization problems taken from two sources. The first set of 35 came from existing benchmark sets available in the literature. The second set came from the 2014 IEEE Single Function optimization competition. The overall results show that CADE has a favorable performance and scalability behaviors when compared to other recent state-of-the-art algorithms. CADEs overall performance ranked at number 1 for each of the two sets of problems. It is suggested that CADEs success across such a broad spectrum of problem types and complexities bodes well for its application to new and novel applications.

An Adaptive Multipopulation Differential Evolution With Dynamic Population Reduction

Developing efficient evolutionary algorithms attracts many researchers due to the existence of optimization problems in numerous real-world applications. A new differential evolution algorithm,

Weaving the social fabric: The past, present and future of optimization problem solving with cultural algorithms

{s}

Leveraged Neighborhood Restructuring in Cultural Algorithms for Solving Real-World Numerical Optimization Problems

TDE-