Adnan Acan

An External Memory Implementation in Ant Colony Optimization

An ant colony optimization algorithm using a library of partial solutions for knowledge incorporation from previous iterations is introduced. Initially, classical ant colony optimization algorithm runs for a small number of iterations and the library of partial solutions is initialized. In this library, variable size solution segments from a number of elite solutions are stored and each segment is associated with its parent’s objective function value. There is no particular distribution of ants in the problem space and the starting point for an ant is the initial point of the segment it starts with. In order to construct a solution, a particular ant retrieves a segment from the library based on its goodness and completes the rest of the solution. Constructed solutions are also used to update the memory. The proposed approach is used for the solution of TSP and QAP for which the obtained results demonstrate that both the speed and solution quality are improved compared to conventional ACO algorithms.

Multiple classifier implementation of a divide-and-conquer approach using appearance-based statistical methods for face recognition

Abstract This paper presents a multiple classifier system for the face recognition problem-based on a novel divide-and-conquer approach using appearance-based statistical methods, namely principal component analysis (PCA), linear discriminant analysis (LDA) and independent component analysis (ICA). A facial image is divided into a number of horizontal segments and the associated local features are extracted using a particular statistical method. Using a simple distance measure and an appropriate classifier combination method, facial images are successfully classified. The standard FERET database and the FERET evaluation methodology are used in all experimental evaluations. Computational and storage space efficiencies and experimental recognition performance of the proposed approach indicate that significant achievements are obtained compared to individual classifiers.

Multi-objective optimization with cross entropy method: Stochastic learning with clustered pareto fronts

This paper presents a novel multiobjective optimization strategy based on the cross entropy method (MOCE). The cross-entropy method (CE) is a stochastic learning algorithm inspired from rare event simulations and proved to be successful in the solution of difficult single objective real-valued optimization problems. The presented work extends the use of cross-entropy method to real-valued multiobjective optimization. For this purpose, parameters of CE search are adapted using the information collected from clustered nondominated solutions on the Pareto front. Comparison with well known multiobjective optimization algorithms on the solution of provably difficult benchmark problem instances demonstrated that CEMO performs at least as good as its competitors.

Chromosome reuse in genetic algorithms

This paper introduces a novel genetic algorithm strategy based on the reuse of chromosomes from previous generations in the creation of offspring individuals. A number of chromosomes of above-average quality, that are not utilized for recombination in the current generation, are inserted into a library called the chromosome library. The main motivation behind the chromosome reuse strategy is to trace some of the untested search directions in the recombination of potentially promising solutions. In the recombination process, chromosomes of current population are combined with the ones in the chromosome library to form a population from which offspring individuals are to be created. Chromosome library is partially updated at the end of each generation and its size is limited by a maximum value. The proposed algorithm is applied to the solution of hard numerical and combinatorial optimization problems. It outperforms the conventional genetic algorithms in all trials.

An external partial permutations memory for ant colony optimization

A novel external memory implementation based on the use of partially complete sequences of solution components from above-average quality individuals over a number of previous iterations is introduced. Elements of such variable-size partial permutation sequences are taken from randomly selected positions of parental individuals and stored in an external memory called the partial permutation memory. Partial permutation sequences are associated with lifetimes together with their parent solutions’ fitness values that are used in retrieving and updating the contents of the memory. When a solution is to be constructed, a partial permutation sequence is retrieved from the memory based on its age and associated fitness value, and the remaining components of the partial solution is completed with an ant colony optimization algorithm. Resulting solutions are also used to update some elements within the memory. The implemented algorithm is used for the solution of a difficult combinatorial optimization problem, namely the quadratic assignment problem, for which significant performance achievements are provided in terms of convergence speed and solution quality.

Probability collectives hybridised with differential evolution for global optimisation

Probability collectives PC is a recent agent-based search framework for function optimisation through optimising parameters of a collection of probability distributions. Differential evolution DE is a successful metaheuristic method particularly for real-parameter global optimisation. This paper presents a hybrid computational model based on a modified PC and DE algorithms for the purpose of improved solutions for real-valued optimisation problems. In the proposed model, PC performs a first phase local search and explores promising search areas through updating parameters of probability distributions over the solution space while DE uses the extracted PC-based knowledge to guide its search with adaptive heuristics. A novel distance-based adaptive mutation scheme is designed within DE to guide the search towards better regions of the solution space. Experimental results reveal that the proposed hybrid algorithm is able to integrate the PCs collective learning methodology and DEs adaptive search strategy effectively to generate improved solutions for difficult problems.

GAACO: A GA + ACO Hybrid for Faster and Better Search Capability

Considering the similarities and characteristics differences between ant colony optimization (ACO) and evolutionary genetic algorithms (GAs), a novel hybrid algorithm combining the search capabilities of the two metaheuristics, for faster and better search capabilities, is introduced. In the GAACO approach, ACO and GAs use identical problem representations and they run in parallel. Migration occurs between the two algorithms whenever any of the them finds an improved potential solution after an iteration. Migration provides further intensification capabilities to both of the algorithms other than their own search mechanisms. In this respect, GAs support ACO by strengthening potential search alternatives for artificial ants and ACO supports GAs by exporting promising potential solutions into its population. The developed algorithm is tested on the solution of two NP-hard combinatorial optimization problems, the obtained results outperform those obtained by both of the individual algorithms when applied alone.

Clonal selection algorithm with operator multiplicity

An artificial immune system using the clonal selection principle with multiple hypermutation operators in its implementation is presented. Mutation operators to be used are identified initially. In every mutation operation, the fitness gain achieved by the employed mutation operator is computed and stored. Accordingly, mutation operators are assigned fitness values based on the fitness improvements they achieve over a number of previous generations. These fitness values are used to determine operator selection probabilities. This approach is used for the solution of a well-known numerical optimization problem, frequency assignment, for which optimal results are achieved in reasonable computation times even for very difficult problem instances.

A genetic algorithm with multiple crossover operators for optimal frequency assignment problem

A genetic algorithm employing multiple crossover operators in its implementation is presented. A set of available crossover operators is established initially and a particular crossover operator is probabilistically selected from this set for recombination. Each crossover operator is assigned a fitness value based on the amount of fitness improvements they achieve over a number of previous generations. Hence, a randomized dynamic operator selection scheme is followed. This approach is used for the solution for which optimal results are achieved in reasonable computation times even for the very difficult problem instances.

A two-stage memory powered Great Deluge algorithm for global optimization

A two-stage memory architecture and search operators exploiting the accumulated experience in memory are maintained within the framework of a Great DeLuge algorithm for real-valued global optimization. The level-based acceptance criterion of the Great DeLuge algorithm is applied for each best solution extracted in a particular iteration. The use of memory-based search supported by effective move operators results in a powerful optimization algorithm. The success of the presented approach is illustrated using three sets of well-known benchmark functions including problems of varying sizes and difficulties. Performance of the presented approach is evaluated and in comparison to well-known algorithms and their published results. Except for a few large-scale optimization problems, experimental evaluations demonstrated that the presented approach performs at least as good as its competitors.