Dr Alan Reynolds
My research work is primarily focused on the application of optimization methods, typically metaheuristics, to a range of different problems. I am particularly interested in the application of such techniques to problems with multiple, competing objectives, through the use of multiobjective metaheuristics that exploit the concept of Pareto dominance of solutions. The effect of modifying this ‘dominance relation’ on the quality of search has been a recurring theme in my publications.
One problem domain that has recurred in my work has been that of data mining and machine learning. I have applied optimization methods to both supervised and unsupervised learning and to feature subset selection.
I have worked with a number of industry collaborators on real world optimization problems, including factory scheduling, call location in mobile phone networks and oil reservoir modelling. Meanwhile, in my spare time I have been attempting to enhance my mathematical knowledge, at present focusing on abstract algebra and number theory.
- Alan P. Reynolds, David W. Corne, Michael J. Chantler (2010) Feature Selection for Multi-Purpose Predictive Models: a Many-Objective Task, Parallel Problem Solving From Nature - PPSN XI. Lecture Notes in Computer Science no. 6239, part 1, pp 384-393.
- Alan P. Reynolds, Beatriz de la Iglesia (2009) A Multi-Objective GRASP for Partial Classification, Soft Computing - A Fusion of Foundations, Methodologies and Applications, vol. 13, no. 3, pp 227-243
- Alan P. Reynolds and Beatriz de la Iglesia (2007) Managing Population Diversity Through the Use of Weighted Objectives and Modified Dominance: An Example from Data Mining, Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Multicriteria Decision Making (MCDM 2007).
- A. P. Reynolds and G. P. McKeown (2007) Construction of factory schedules using reverse simulation, European Journal of Operational Research, vol. 179, no. 3, pp 656-676.
- A. P. Reynolds, J. L. Dicks, I. N. Roberts, J. J. Wesselink, B. de la Iglesia, V. Robert, T. Boekhout and V. J. Rayward-Smith (2003) Algorithms for Identification Key Generation and Optimization with Application to Yeast Identification, Applications of Evolutionary Computing: EvoBIO 2003. Lecture Notes in Computer Science no. 2611, pp 107-118.
My undergraduate years were spent at the University of Cambridge, studying mathematics for three years to obtain my degree and for a fourth to study ‘Part III’. While I maintain an interest in mathematics, I converted to computer science via Cambridge University’s diploma.
Leaving Cambridge, my research career started at the University of East Anglia (UEA) where I obtained my Ph.D. applying optimization algorithms to both real-world and more academic scheduling problems. A number of research contracts followed, on a diverse range of topics, from the identification of yeast to the clustering of rules.
After a number of years as research associate and senior research associate at UEA, I left to work with Prof. David Corne on an interesting industrial optimization project at Heriot-Watt University. At Heriot-Watt, I continue to apply my expertise to both the fundamentals of optimization algorithm design and the application of optimization techniques to industrial problems.