Andrew Tuson

Adapting operator settings in genetic algorithms

In the majority of genetic algorithm implementations, the operator settings are fixed throughout a given run. However, it has been argued that these settings should vary over the course of a genetic algorithm runso as to account for changes in the ability of the operators to produce children of increased fitness. This paper describes an investigation into this question. The effect upon genetic algorithm performance of two adaptation methods upon both well-studied theoretical problems and a hard problem from operations research, the flowshop sequencing problem, are therefore examined. The results obtained indicate that the applicability of operator adaptation is dependent upon three basic assumptions being satisfied by the problem being tackled.

Evolving Musical Harmonisation

We describe a series of experiments in generating traditional musical harmony using Genetic Algorithms. We discuss some problems which are specific to the musical domain, and conclude that a GA with no notion of metalevel control of the reasoning process is unlikely to solve the harmonisation problem well.

Cost Based Operator Rate Adaption: An Investigation

In the vast majority of genetic algorithm implementations, the operator probabilities are fixed throughout a given run. However, it may be useful to adjust these probabilities during the run, according to the ability of the operators to produce children of increased fitness. Cost Based Operator Rate Adaptation (COBRA) periodically re-ranks operator probabilities according to a measure of operator performance. The effect upon genetic algorithm performance of COBRA upon both well-studied theoretical and practical problems is examined.

A Voxel-Based Representation for the Evolutionary Shape Optimisation of a Simplified Beam: A Case-Study of a Problem-Centred Approach to Genetic Operator Design

This paper examines a voxel (N-dimcnsional pixel) based representation for shape optimisation problems, and shows that although a basic genetic algorithm performed poorly on a simplified beam design problem, the use of three domain specific operators improved performance greatly. Additionally, the use of a ‘directed smoothing’ operator that preferentially adds material to high stress areas was examined and found to assist evolutionary search. This paper demonstrates how domain knowledge and an understanding of how genetic algorithms work can be used to inform the design of suitable operators.

Co-evolution of Operator Settings in Genetic Algorithms

Typical genetic algorithm implementations use operator settings that are fixed throughout a given run. Varying these settings is known to improve performance — the problem is knowing how to vary them. One approach is to encode the operator settings into each member of the GA population, and allow them to evolve. This paper describes an empirical investigation into the effect of co-evolving operator settings, for some common problems in the genetic algorithms field. The results obtained indicate that the problem representation, and the choice of operators on the encoded operator settings are important for useful adaptation.

An Evolutionary/Meta-Heuristic Approach to Emergency Resource Redistribution in the Developing World

The problem of logistics and resource management in disease control projects in the developing world can hardly be understated. One example is the occurance of regional imbalances in supply. A prototype system, based upon evolutionary and ‘meta-heuristic’ optimisation techniques is described that recommends a plan for the redistribution of available resources to minimise shortages. Evaluation of the system on data from real world situations indicated that the generation of good, feasible redistribution plans is possible even on large datasets. Comparison of the optimisers showed that evolutionary techniques perform poorly on this problem compared to stochastic hill climbing.

Directing the Search of Evolutionary and Neighbourhood-Search Optimisers for the Flowshop Sequencing Problem with an Idle-Time Heuristic

This paper presents a heuristic for directing the neighbourhood (mutation operator) of stochastic optimisers, such as evolutionary algorithms, so to improve performance for the flowshop sequencing problem. Based on idle time, the heuristic works on the assumption that jobs that have to wait a relatively long time between machines are in an unsuitable position in the schedule and should be moved. The results presented here show that the heuristic improves performance, especially for problems with a large number of jobs. In addition the effectiveness of the heuristic and search in general was found to depend upon the neighbourhood structure in a consistent fashion across optimisers.

A Voxel Based Approach to Evolutionary Shape Optimisation

Shape optimisation is a hard problem from the field of Mechanical Engineering with the potential for significant cost savings if successfully performed. In the past, evolutionary optimisation approaches have proved successful. In those studies, a form of the shape was assumed, and its parameters optimised. An alternative is a voxel (N-dimensional pixel) based representation, which makes no such assumptions about the form of the solution, and allows the user to add domain knowledge if desired. This paper outlines a preliminary investigation into this approach and shows that the objections to this approach in the literature can be overcome if care is taken over the design of the operators.

Evolving Protein Similarity Scoring Matrices Using Differential Evolution

The problem of protein structure prediction remains one of the major unsolved problems in structural biochemistry. The most successful method to date for predicting protein tertiary (3-dimensional) structure from the primary sequence data (of amino acids) is homology modelling based on alignment with similar sequences of known structure. The premier component utilised in this process is a scoring matrix that determines how similar one protein is to another.