Requisites: Pre-requisite: programming skills equivalent to level 9

The module introduces heuristic methods for solving complex optimisation problems such as routing, scheduling, timetabling that occur in the real world and cannot be solving in practical timeframes by mathematical techniques.It will first describe why some problems are hard to solve, and then go on to look in depth at a range of modern heuristic search methods such as Evolutionary Algorithms (including evolution strategies and other meta-heuristic search techniques) as local search methods such as tabu-search, simulated annealing.The module will describe appropriate experimental methodologies for testing and analysing stochastic search algorithms and introduce basic statistics for analysing their performance. Students will have the opportunity to put the methods into practice during lab sessions and use their knowledge to solve a real world problem in the coursework.

Upon completion of this module you will be able to

LO1: Assess the applicability of different optimisation algorithms to different problem domains.

LO2: Critically assess the role played by algorithm operators in designing an algorithm.

LO3: Specify, customise and use appropriate algorithms for particular practical problems.

LO4: Apply an appropriate scientific approach to testing and analysing emergent computing algorithms and be able to use simple statistics to evaluate the results.

LO5: Present the results in an appropriate form and draw appropriate conclusions.