Effective learning hyper-heuristics for the course timetabling problem

Soria-Alcaraz, Jorge A.; Ochoa, Gabriela; Swan, Jerry; Carpío, Martín; Soberanes, Héctor José Puga; Burke, Edmund K.

doi:10.1016/j.ejor.2014.03.046

Cited by 73 publications

(55 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Each LLH operates a small shift in current state of solution. 3 Shift_LessSituratedDay Day Shift Semi -R LLH 4 Shift_DayConstToLessSituDay Day Shift Semi -R LLH 5 Shift_WithDayConstImprov Period Shift Progressive LLH 6 Swap_InDays Period Swap Semi -R LLH 7 Shift_RandDayImprovment Day Shift Progressive LLH 8 Swap_InColumn Period Swap Progressive LLH 9 Shift_NeighboringPeroid Period Shift Random LLH 10 Swap_InRows Period Swap Progressive LLH 11 Shift_Dispersions_of _Exam Day Shift Progressive Table 2 exhibits the classification of Low Level Heuristics (LLHs) designed and developed in this research project, however last four LLHs are described in this article. The some LLHs are laying either in Shift or Swap functionality.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Formulation of low level heuristics

Shaikh

Raziq

Muhammad

2014

IJBAS

View full text Add to dashboard Cite

The curricula scheduling is very significant and largely studied problem in academia. The desired solution calculatedly assembles the academic events over the carefully designed layout considering several predefined interlinked constraints. The contemporary research for solving scheduling constraints is inclined to raise the degree of generality, so that a wide range of identical problems may be addressed. The hyper-heuristic is such a state-of-the-art solving technique which stands on multi-layered framework. The top layer usually consists of classic algorithm for managing the operators on down-layers, and the same is occasionally assisted by machine learning or similar techniques. This research article examines the performance of the small group of bespoke low level heuristics. These LLHs are operated by hyperheuristic to address the specific scheduling constraints. The set of heuristics are divided into a range of subgroups including timescale category which contain two subsets Day and Period. The utility group which contains two patterns named Shift and Swap techniques, while the third category encircles three more subgroups of Random or Sami-Random and Progressive.

show abstract

Section: Methodsmentioning

confidence: 99%

“…Contemporary research direction in TTP is tending to raise the level of generality by state-of-art techniques so that a range of instances can be tackled. Hyper-heuristic is one of such modern-day techniques that largely shape such idea [7], [8].…”

Section: Related Workmentioning

confidence: 99%

Formulation of low level heuristics

Shaikh

Raziq

Muhammad

2014

IJBAS

View full text Add to dashboard Cite

show abstract

“…The local search operator is in charge of improving solutions, while the perturbation operator moves to other regions in the search space to escape from local optimal solutions. One of the motivations behind including ILS in this investigation as an alternative search to evolutionary search, was its simple implementation and promising results provided by hyper-heuristic implementations in other domains [10,40,44].…”

Section: Iterated Local Search Hyper-heuristicmentioning

confidence: 99%

Grammar-based generation of variable-selection heuristics for constraint satisfaction problems

Sosa-Ascencio

Ochoa

Terashima-Marín

et al. 2015

Genet Program Evolvable Mach

Self Cite

View full text Add to dashboard Cite

We propose a grammar-based genetic programming framework that generates variable-selection heuristics for solving constraint satisfaction problems. This approach can be considered as a generation hyper-heuristic. A grammar to express heuristics is extracted from successful human-designed variable-selection heuristics. The search is performed on the derivation sequences of this grammar using a strongly typed genetic programming framework. The approach brings two innovations to grammar-based hyper-heuristics in this domain: the incorporation of if-then-else rules to the function set, and the implementation of overloaded functions capable of handling different input dimensionality. Moreover, the heuristic search space is explored using not only evolutionary search, but also two alternative simpler strategies, namely, iterated local search and parallel hill climbing. We tested our approach on synthetic and real-world instances. The newly generated heuristics have an improved performance when compared against human-designed heuristics. Our results suggest that the constrained search space imposed by the proposed grammar is the main factor in the generation of good heuristics. However, to generate more general heuristics, the composition of the training set and the search methodology played an important role. We found that increasing the variability of the training set improved the generality of the evolved heuristics, and the evolutionary search strategy produced slightly better results.

show abstract

“…It is also known that different components can perform well at different points in the search (see e.g. [37]), which is particularly important in the case of dynamic environments [27]. One method of composition for components is to use the 'Composite' Design Pattern [13], i.e.…”

Section: The Design Of Haikumentioning

confidence: 99%

Haiku - a Scala Combinator Toolkit for Semi-automated Composition of Metaheuristics

Kocsis

Brownlee

Swan

et al. 2015

Search-Based Software Engineering

Self Cite

View full text Add to dashboard Cite

Abstract. There is an emerging trend towards the automated design of metaheuristics at the software component level. In principle, metaheuristics have a relatively clean decomposition, where well-known frameworks such as ILS and EA are parameterised by variant components for acceptance, perturbation etc. Automated generation of these frameworks is not so simple in practice, since the coupling between components may be implementation specific. Compositionality is the ability to freely express a space of designs 'bottom up' in terms of elementary components: previous work in this area has used combinators, a modular and functional approach to componentisation arising from foundational Computer Science. In this article, we describe Haiku, a combinator toolkit written in the Scala language, which builds upon previous work to further automate the process by automatically composing the external dependencies of components. We provide examples of use and give a case study in which a programatically-generated heuristic is applied to the Travelling Salesman Problem within an Evolutionary Strategies framework.

show abstract

Effective learning hyper-heuristics for the course timetabling problem

Cited by 73 publications

References 37 publications

Formulation of low level heuristics

Formulation of low level heuristics

Grammar-based generation of variable-selection heuristics for constraint satisfaction problems

Haiku - a Scala Combinator Toolkit for Semi-automated Composition of Metaheuristics

Contact Info

Product

Resources

About