Abdellah Salhi

Multiobjective memetic algorithm based on decomposition

In recent years, hybridization of multi-objective evolutionary algorithms (MOEAs) with traditional mathematical programming techniques have received significant attention in the field of evolutionary computing (EC). The use of multiple strategies with self-adaptation manners can further improve the algorithmic performances of decomposition-based evolutionary algorithms. In this paper, we propose a new multiobjective memetic algorithm based on the decomposition approach and the particle swarm optimization (PSO) algorithm. For brevity, we refer to our developed approach as MOEA/D-DE+PSO. In our proposed methodology, PSO acts as a local search engine and differential evolution works as the main search operator in the whole process of optimization. PSO updates the position of its solution with the help of the best information on itself and its neighboring solution. The experimental results produced by our developed memtic algorithm are more promising than those of the simple MOEA/D algorithm, on most test problems. Results on the sensitivity of the suggested algorithm to key parameters such as population size, neighborhood size and maximum number of solutions to be altered for a given subproblem in the decomposition process are also included.

A decomposition-based hybrid multiobjective evolutionary algorithm with dynamic resource allocation

Different crossover operators suit different problems. It is, therefore, potentially problematic to chose the ideal crossover operator in an evolutionary optimization scheme. Using multiple crossover operators could be an effective way to address this issue. This paper reports on the implementation of this idea, i.e. the use of two crossover operators in a decomposition-based multi-objective evolutionary algorithm, but not simultaneously. After each cycle, the operator which has helped produce the better offspring is rewarded. This means that the overall algorithm uses a dynamic resource allocation to reward the better of the crossover operators in the optimization process. The operators used are the Simplex Crossover operator (SPX) and the Center of Mass Crossover operator (CMX). We report experimental results that show that this innovative use of two crossover operators improves the algorithm performance on standard test problems. Results on the sensitivity of the suggested algorithm to key parameters such as population size, neighborhood size and maximum number of solutions to be altered for a given subproblem in the the decomposition process are also included.

An estimation of distribution algorithm with guided mutation for a complex flow shop scheduling problem

An Estimation of Distribution Algorithm (EDA) is proposed toapproach the Hybrid Flow Shop with Sequence Dependent Setup Times and Uniform Machines in parallel (HFS-SDST-UM) problem. The latter motivated by the needs of a real world company. The proposed EDA implements a fairly new mechanism to improve the search of more traditional EDAs. This is the Guided Mutation (GM). EDA-GM generates new solutions by using the information from a probability model, as all EDAs, and the local information from a good known solution. The approach is tested on several instances of HFS-SDST-UM and compared with adaptations of meta-heuristics designed for very similarproblems. Encouraging results are reported.

Parallel implementation of a genetic-programming based tool for symbolic regression

We report on a parallel implementation of a tool for symbolic regression, the algorithmic mechanism of which is based on genetic programming, and communication is handled using MPI. The implementation relies on a random islands model (RIM), which combines both the conventional islands model where migration of individuals between islands occurs periodically and niching where no migration takes place. The system was designed so that the algorithm is synergistic with parallel/distributed architectures, and works to make use of processor time and minimum use of network bandwidth without complicating the sequential algorithm significantly. Results on an IBM SP2 are included.

A Plant Propagation Algorithm for Constrained Engineering Optimisation Problems

Optimisation problems arising in industry are some of the hardest, often because of the tight specifications of the products involved. They are almost invariably constrained and they involve highly nonlinear, and non-convex functions both in the objective and in the constraints. It is also often the case that the solutions required must be of high quality and obtained in realistic times. Although there are already a number of well performing optimisation algorithms for such problems, here we consider the novel Plant Propagation Algorithm (PPA) which on continuous problems seems to be very competitive. It is presented in a modified form to handle a selection of problems of interest. Comparative results obtained with PPA and state-of-the-art optimisation algorithms of the Nature-inspired type are presented and discussed. On this selection of problems, PPA is found to be as good as and in some cases superior to these algorithms.

Mathematical programming and the sensitivity of multi-criteria decisions

We report on the current state of a project whose aim is to implement a framework for sensitivity analysis in Multi-Criteria Decision Making (MCDM). The framework is largely based on mathematical programming. Due to the potentially large number and nature of the mathematical programmes, it is far from trivial to provide solutions to all of them in acceptable computing times. The challenge is even greater when we recognize that much decision analysis is performed in the context of decision conferences where any sensitivity analysis needs to be conducted in near real time (preferably) on a PC. We present a parallel processing approach to this challenge and point to some of the difficulties still to be resolved. Preliminary results obtained on a network of transputers are discussed.

Tailoring hyper-heuristics to specific instances of a scheduling problem using affinity and competence functions

Hyper-heuristics are high level heuristics which coordinate lower level ones to solve a given problem. Low level heuristics, however, are not all as competent/good as each other at solving the given problem and some do not work together as well as others. Hence the idea of measuring how good they are (competence) at solving the problem and how well they work together (their affinity). Models of the affinity and competence properties are suggested and evaluated using previous information on the performance of the simple low level heuristics. The resulting model values are used to improve the performance of the hyper-heuristic by tailoring it not only to the specific problem but the specific instance being solved. The test case is a hard combinatorial problem, namely the Hybrid Flow Shop scheduling problem. Numerical results on randomly generated as well as real-world instances are included.

Sensitivity analysis in statistical decision theory: A decision analytic view

Sensitivity analysis provides a way to mitigate traditional criticisms of Bayesian statistical decision theory, concerning dependence on subjective inputs. We suggest a general framework for sensitivity analysis allowing for perturbations in both the utility function and the prior distribution. Perturbations are constrained to classes modelling imprecision in judgements The framework discards first definitely bad alternatives; then, identifies alternatives that may share optimality with a current one; and, finally, detects least changes in the inputs leading to changes in ranking. The associated computational problems and their implementation are discussed.

A Game Theory-Based Multi-Agent System for Expensive Optimisation Problems

This paper is concerned with the development of a novel approach to solve expensive optimisation problems. The approach relies on game theory and a multi-agent framework in which a number of existing algorithms, cast as agents, are deployed with the aim to solve the problem in hand as efficiently as possible. The key factor for the success of this approach is a dynamic resource allocation biased toward promising algorithms on the given problem. This is achieved by allowing the agents to play a cooperative-competitive game the outcomes of which will be used to decide which algorithms, if any, will drop out of the list of solver-agents and which will remain in use. A successful implementation of this framework will result in the most suited algorithm(s) for the given problem being predominantly used on the available computing platform. In other words it guarantees the best use of the resources both algorithms and hardware with the by-product being the best approximate solution for the problem given the available resources. GTMAS is tested on a standard collection of TSP problems. The results are included.

A Seed-Based Plant Propagation Algorithm: The Feeding Station Model

The seasonal production of fruit and seeds is akin to opening a feeding station, such as a restaurant. Agents coming to feed on the fruit are like customers attending the restaurant; they arrive at a certain rate and get served at a certain rate following some appropriate processes. The same applies to birds and animals visiting and feeding on ripe fruit produced by plants such as the strawberry plant. This phenomenon underpins the seed dispersion of the plants. Modelling it as a queuing process results in a seed-based search/optimisation algorithm. This variant of the Plant Propagation Algorithm is described, analysed, tested on nontrivial problems, and compared with well established algorithms. The results are included.