Streamlined constraint reasoning : an automated approach from high level constraint specifications
MetadataShow full item record
Constraint Programming (CP) is a powerful technique for solving large-scale combinatorial (optimisation) problems. Solving a problem proceeds in two distinct phases: modelling and solving. Effective modelling has a huge impact on the performance of the solving process. Even with the advance of modern automated modelling tools, search spaces involved can be so vast that problems can still be difficult to solve. To further constrain the model a more aggressive step that can be taken is the addition of streamliner constraints, which are not guaranteed to be sound but are designed to focus effort on a highly restricted but promising portion of the search space. Previously, producing effective streamlined models was a manual, difficult and time-consuming task. This thesis presents a completely automated process to the generation, search and selection of streamliner portfolios to produce a substantial reduction in search effort across a diverse range of problems. First, we propose a method for the generation and evaluation of streamliner conjectures automatically from the type structure present in an Essence specification. Second, the possible streamliner combinations are structured into a lattice and a multi-objective search method for searching the lattice of combinations and building a portfolio of streamliner combinations is defined. Third, the problem of "Streamliner Selection" is introduced which deals with selecting from the portfolio an effective streamliner for an unseen instance. The work is evaluated by presenting two sets of experiments on a variety of problem classes. Lastly, we explore the effect of model selection in the context of streamlined specifications and discuss the process of streamlining for Constrained Optimization Problems.
Thesis, PhD Doctor of Philosophy
Items in the St Andrews Research Repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Showing items related by title, author, creator and subject.
Using constraints to improve generalisation and training of feedforward neural networks : constraint based decomposition and complex backpropagation Draghici, Sorin (University of St Andrews, 1996) - ThesisNeural networks can be analysed from two points of view: training and generalisation. The training is characterised by a trade-off between the 'goodness' of the training algorithm itself (speed, reliability, guaranteed ...
Environmental constraints on the foraging behaviour, spatial usage and population sizes of albatrosses Wakefield, Ewan D. (University of St Andrews, 2009) - ThesisSatellite-tracking of wide-ranging, apex marine predators, combined with remote-sensing, can be used to test ecological hypotheses and to estimate spatial abundance. I used this approach to quantify the habitat usage of ...
Nightingale, Peter (University of St Andrews, 2007-11-30) - ThesisConstraint satisfaction is a very well studied and fundamental artificial intelligence technique. Various forms of knowledge can be represented with constraints, and reasoning techniques from disparate ﬁelds can be ...