Jorge A. Soria-Alcaraz

Effective learning hyper-heuristics for the course timetabling problem

Course timetabling is an important and recurring administrative activity in most educational institutions. This article combines a general modeling methodology with effective learning hyper-heuristics to solve this problem. The proposed hyper-heuristics are based on an iterated local search procedure that autonomously combines a set of move operators. Two types of learning for operator selection are contrasted: a static (offline) approach, with a clear distinction between training and execution phases; and a dynamic approach that learns on the fly. The resulting algorithms are tested over the set of real-world instances collected by the first and second International Timetabling competitions. The dynamic scheme statistically outperforms the static counterpart, and produces competitive results when compared to the state-of-the-art, even producing a new best-known solution. Importantly, our study illustrates that algorithms with increased autonomy and generality can outperform human designed problem-specific algorithms.

Iterated local search using an add and delete hyper-heuristic for university course timetabling

Graphical abstractDisplay Omitted HighlightsAdd and delete operations are encoded as a list/string of integers (ADL).An effective hyper-heuristic approach operating with ADLs is proposed.Low level heuristics perform search over the space of feasible solutions.Proposed approach produces new best solutions to some instances.Proposed approach achieves generality across two variants of the timetabling problem. Hyper-heuristics are (meta-)heuristics that operate at a higher level to choose or generate a set of low-level (meta-)heuristics in an attempt of solve difficult optimization problems. Iterated local search (ILS) is a well-known approach for discrete optimization, combining perturbation and hill-climbing within an iterative framework. In this study, we introduce an ILS approach, strengthened by a hyper-heuristic which generates heuristics based on a fixed number of add and delete operations. The performance of the proposed hyper-heuristic is tested across two different problem domains using real world benchmark of course timetabling instances from the second International Timetabling Competition Tracks 2 and 3. The results show that mixing add and delete operations within an ILS framework yields an effective hyper-heuristic approach.

Comparison of PSO and DE for Training Neural Networks

The use of computational resources required for Feed-Forward Artificial Neural Network (FFANN) training phase by means of classical techniques such as the back propagation learning rule can be prohibitive in some applications. A good training phase is needed for a high performance of a neural network. In searching for alternative methods for training phase of FFANN, some metaheuristic techniques have been used to do this task. This paper compares the performance of Particle Swarm Optimization (PSO) and Differential Evolution (DE) as training methods for FFANN under several well-known pattern recognition instances.

Evolving and reusing Bin Packing heuristic through Grammatical Differential Evolution

The Bin Packing Problem is a classic optimization problem, over the years many heuristics have been developed to obtain better results. There are many approaches to generating heuristics automatically, those approaches are based Genetic Programming, but the heuristics generated sometimes can not be applied to the problem. Recently in the Artificial Intelligence field, the Grammar Evolution approach emerged, which generated expressions like the generated by Genetic Programming; these algorithms evolve into a grammar based on the Backus Naur Form. In the present work we show a Grammar Evolution based on Differential Evolution, which automatically generated heuristics for the Bin Packing Problem instances. Those heuristics generated by the Grammar Evolution are like the Best-Fit heuristic which was designed by humans. The works goal is to prove that is feasible to use the Grammar Evolution to automatically generate and reusing heuristics which have at least the same performance than the best generated by humans, we also propose a Grammar to improve the results obtained for a Grammar based on Genetic Programming.

A methodology for determining an effective subset of heuristics in selection hyper-heuristics

We address the important step of determining an effective subset of heuristics in selection hyper-heuristics. Little attention has been devoted to this in the literature, and the decision is left at the discretion of the investigator. The performance of a hyper-heuristic depends on the quality and size of the heuristic pool. Using more than one heuristic is generally advantageous, however, an unnecessary large pool can decrease the performance of adaptive approaches. Our goal is to bring methodological rigour to this step. The proposed methodology uses non-parametric statistics and fitness landscape measurements from an available set of heuristics and benchmark instances, in order to produce a compact subset of effective heuristics for the underlying problem. We also propose a new iterated local search hyper-heuristic using multi-armed bandits coupled with a change detection mechanism. The methodology is tested on two real-world optimization problems: course timetabling and vehicle routing. The proposed hyper-heuristic with a compact heuristic pool, outperforms state-of-the-art hyper-heuristics and competes with problem-specific methods in course timetabling, even producing new best-known solutions in 5 out of the 24 studied instances.

Improving the Bin Packing Heuristic through Grammatical Evolution Based on Swarm Intelligence

In recent years Grammatical Evolution (GE) has been used as a representation of Genetic Programming (GP) which has been applied to many optimization problems such as symbolic regression, classification, Boolean functions, constructed problems, and algorithmic problems. GE can use a diversity of searching strategies including Swarm Intelligence (SI). Particle Swarm Optimisation (PSO) is an algorithm of SI that has two main problems: premature convergence and poor diversity. Particle Evolutionary Swarm Optimization (PESO) is a recent and novel algorithm which is also part of SI. PESO uses two perturbations to avoid PSO’s problems. In this paper we propose using PESO and PSO in the frame of GE as strategies to generate heuristics that solve the Bin Packing Problem (BPP); it is possible however to apply this methodology to other kinds of problems using another Grammar designed for that problem. A comparison between PESO, PSO, and BPP’s heuristics is performed through the nonparametric Friedman test. The main contribution of this paper is proposing a Grammar to generate online and offline heuristics depending on the test instance trying to improve the heuristics generated by other grammars and humans; it also proposes a way to implement different algorithms as search strategies in GE like PESO to obtain better results than those obtained by PSO.

Several Strategies to Improve the Performance of Hyperheuristics for Academic Timetabling Design Problem

The Educational timetabling problem is a common and hard problem inside every educative institution, this problem tries to coordinate Students, Teachers, Classrooms and Timeslots under certain constrains that dependent in many cases the policies of each educational institution. The idea behind hyper-heuristics is to discover some combination of straightforward heuristics to solve a wide range of problems. This paper presents several strategies to improve a GA-based method that produces general hyper heuristics for the educational timetabling design problem using API-Carpio methodology. The GA uses static-length representation, witch involves the complete encoding of a solution algorithm capable to solve schedule design instances. This hyper-heuristic is achieved by learning and testing phases using real instances from Leon Institute of Technology (LIT) producing encouraging results for most of the instances. Finally we analyze the quality of our hyper-heuristic in the context of real Academic Timetabling Design problem against past Investigations.

Increase Methodology of Design of Course Timetabling Problem for Students, Classrooms, and Teachers

The aim of the Course Timetabling problem is to ensure that all the students take their required classes and adhere to resources that are available in the school. The set of constraints those must be considered in the design of timetabling involves students, teachers, and classrooms. In the state of the art are different methodologies of design for Course Timetabling problem, in this paper we extend the proposal from Soria in 2013, in which they consider variables of students and classrooms, with four set of generic structures. This paper uses Soria’s methodology to adding two more generic structures considering teacher restriction. We show an application of some different Metaheuristics using this methodology. Finally, we apply nonparametric test Wilcoxon signed-rank with the aim to find which metaheuristic algorithm shows a better performance in terms of quality.

Evolvability Metric Estimation by a Parallel Perceptron for On-Line Selection Hyper-Heuristics

Online hyper-heuristic selection is a novel and powerful approach to solving complex problems. This approach dynamically selects, based on the state of a given solution, the most promising operator (from a pool of operators) to continue the search process. The dynamic selection is usually based on the analysis of the latest applications of a given operator during actual execution, estimating the potential success of the operator at the current solution state. The estimation can be made by evolvability metrics. Calculating an evolvability metric is computationally expensive since it requires the generation and evaluation of a neighborhood of solutions. This paper aims to estimate the potential success of an operator for a given solution state by using a pre-trained neural network; known as a parallel perceptron. The proposal accelerates the online selection process, allowing us to achieve better performance than hyper-heuristic models, which directly use evolvability functions.

Combining Mutation and Recombination to Improve a Distributed Model of Adaptive Operator Selection

We present evidence indicating that adding a crossover island greatly improves the performance of a Dynamic Island Model for Adaptive Operator Selection. Two combinatorial optimisation problems are considered: the Onemax benchmark, to prove the concept; and a real-world formulation of the course timetabling problem to test practical relevance. Crossover is added to the recently proposed dynamic island adaptive model for operator selection which considered mutation only. When comparing the models with and without a recombination, we found that having a crossover island significantly improves the performance. Our experiments also provide compelling evidence of the dynamic role of crossover during search: it is a useful operator across the whole search process. The idea of combining different type of operators in a distributed adaptive search model is worth further investigation.