David A. Pelta

Optimization in dynamic environments: a survey on problems, methods and measures

This paper provides a survey of the research done on optimization in dynamic environments over the past decade. We show an analysis of the most commonly used problems, methods and measures together with the newer approaches and trends, as well as their interrelations and common ideas. The survey is supported by a public web repository, located at http://www.dynamic-optimization.org where the collected bibliography is manually organized and tagged according to different categories.

Nature Inspired Cooperative Strategies for Optimization (NICSO 2011)

Extending the ABC-Miner Bayesian Classification Algorithm.- A Multiple Pheromone Ant Clustering Algorithm.- An Island Memetic Differential Evolution Algorithm for the Feature Selection Problem.- Using a Scouting Predator-Prey Optimizer to Train Support Vector Machines with non PSD Kernels.- Response Surfaces with Discounted Information for Global Optima Tracking in Dynamic Environments.- Fitness based Self Adaptive Differential.- Adaptation schemes and dynamic optimization problems: a basic study on the Adaptive Hill Climbing Memetic Algorithm.- Using base position errors in an entropy-based evaluation function for the study of genetic code adaptability.- An Adaptive Multi-Crossover Population Algorithm for Solving Routing Problems.- Corner Based Many-Objective Optimization.- Escaping Local Optima via Parallelization and.- An Improved Genetic Based Keyword Extraction Technique.- Part-of-Speech Tagging Using Evolutionary Computation.- A Cooperative approach using ants and bees for the graph coloring problem.- Artificial Bee Colony Training of Neural Networks.- Nonlinar optimization in landscapes with planar regions.- Optimizing Neighbourhood Distances for a Variant of Fully-Informed Particle Swarm Algorithm.- Meta Morphic Particle Swarm Optimization.- Empirical study of computational intelligence strategies for biochemical systems modelling.- Metachronal waves in Cellular Automata: Cilia-like manipulation in actuator arrays.- Team of A-Teams Approach for Vehicle Routing Problem with Time Windows.- Self-adaptable Group Formation of Reconfigurable Agents in Dynamic Environments.- A Choice Function Hyper-Heuristic for the Winner Determination Problem.- Automatic Generation of Heuristics for Constraint Satisfaction Problems.- Branching Schemes and Variable Ordering Heuristics for Constraint Satisfaction Problems: Is there Something to Learn.- Nash Equilibria Detection for Discrete-time Generalized Cournot Dynamic Oligopolies.

The 2009 Simulated Car Racing Championship

In this paper, we overview the 2009 Simulated Car Racing Championship-an event comprising three competitions held in association with the 2009 IEEE Congress on Evolutionary Computation (CEC), the 2009 ACM Genetic and Evolutionary Computation Conference (GECCO), and the 2009 IEEE Symposium on Computational Intelligence and Games (CIG). First, we describe the competition regulations and the software framework. Then, the five best teams describe the methods of computational intelligence they used to develop their drivers and the lessons they learned from the participation in the championship. The organizers provide short summaries of the other competitors. Finally, we summarize the championship results, followed by a discussion about what the organizers learned about 1) the development of high-performing car racing controllers and 2) the organization of scientific competitions.

A simple and fast heuristic for protein structure comparison

BackgroundProtein structure comparison is a key problem in bioinformatics. There exist several methods for doing protein comparison, being the solution of the Maximum Contact Map Overlap problem (MAX-CMO) one of the alternatives available. Although this problem may be solved using exact algorithms, researchers require approximate algorithms that obtain good quality solutions using less computational resources than the formers.ResultsWe propose a variable neighborhood search metaheuristic for solving MAX-CMO. We analyze this strategy in two aspects: 1) from an optimization point of view the strategy is tested on two different datasets, obtaining an error of 3.5%(over 2702 pairs) and 1.7% (over 161 pairs) with respect to optimal values; thus leading to high accurate solutions in a simpler and less expensive way than exact algorithms; 2) in terms of protein structure classification, we conduct experiments on three datasets and show that is feasible to detect structural similarities at SCOPs family and CATHs architecture levels using normalized overlap values. Some limitations and the role of normalization are outlined for doing classification at SCOPs fold level.ConclusionWe designed, implemented and tested.a new tool for solving MAX-CMO, based on a well-known metaheuristic technique. The good balance between solutions quality and computational effort makes it a valuable tool. Moreover, to the best of our knowledge, this is the first time the MAX-CMO measure is tested at SCOPs fold and CATHs architecture levels with encouraging results.Software is available for download at http://modo.ugr.es/jrgonzalez/msvns4maxcmo.

A modular parametric architecture for the TORCS racing engine

This paper presents our approach to TORCS Car Racing Competition 2009, it is based on a complete modular architecture capable of driving automatically a car along a track with or without oppents. The architecture is composed of five simple modules being each one responsible for a basic aspect of car driving. The modules control gear shiftings, steer movements and pedals positions by using of simple functions meanwhile the allowed speed in a certain track segment is managed by a simple TSK fuzzy system.

Using memory and fuzzy rules in a co-operative multi-thread strategy for optimization

In this article, we analyze a co-operative multi-thread search-based optimization strategy, where each solver thread represents a different optimization algorithm (or the same one with different settings), and they are all controlled by a centralized co-ordinator. We also propose the use of memory to keep track of both the state of the individual threads and the obtained solutions. Based on this memory, a very simple fuzzy rule base is used to control the system behavior. We also present the results of three computational experiments. The first of these checks the strategy by comparing it with an independent search strategy and a sequential algorithm, and the superiority of the co-operative scheme is confirmed. The second analyzes how definition of the threads affects the quality of the results, and the importance of there being a balanced set between intensification and diversification is corroborated. The third explores the use of memory with two different fuzzy rules, and the results indicate that the best combination is to use memory together with two rules (solver dependent and solver independent ones) (although this combination should not be activated at the beginning of the search in order to avoid premature convergence).

Application of fuzzy optimization to diet problems in Argentinean farms

Abstract The problem of designing diets for cattle in an Argentinean farm is addressed in this paper. Usually the livestock is not confined, so it is impossible to assess the amount of food each animal will eat. Therefore it makes no sense to design diets verifying the nutritional requirements exactly. It is more valuable to allow for constraint violations, i.e., fuzzy constraints which in turn may lead to cheaper diets. Under this scenario the problem is modelled as a fuzzy linear programming (FLP) one and then solved by using a decision support systems (DSS) named SACRA (a Spanish acronym for support system for the construction of cattle diets) that the authors have specifically developed for this problem. SACRA is based on PROBO, an already experimented DSS also developed by the authors, and it is highly friendship, interactive and does not require any knowledge about FLP. The tests carried out with SACRA have shown a high level of satisfaction from the side of the decision makers.

A fuzzy sets based generalization of contact maps for the overlap of protein structures

The comparison of protein structures is an important problem in bioinformatics. As a protein biological role is derived from its three-dimensional native state, the comparison of a new protein structure (with unknown function) with other protein structures (with known biological activity) can shed light into the biological role of the former. Consequently, advances in the comparison (and clustering) of proteins according to their three-dimensional configurations might also have an impact on drug discovery and other biomedical research that relies on understanding the inter-relations between structure and function in proteins. The contributions described in this paper are: Firstly, we propose a generalization of the maximum contact map overlap problem (MAX-CMO) by means of fuzzy sets and systems. The MAX-CMO is a model for protein structure comparison. In our new model, namedgeneralized maximum fuzzy contact map overlap (GMAX-FCMO), a contact map is defined by means of one (or more) fuzzy thresholds and one (or more) membership functions. The advantages and limitations of our new model are discussed. Secondly, we show how a fuzzy sets based metaheuristic can be used to compute protein similarities based on the new model. Finally, we compute the protein structure similarity of real-world proteins and show how our new model correctly measures their (di)similarity.

A Fuzzy Valuation-Based Local Search Framework for Combinatorial Problems

AbstractA novel local search method is presented. One of the new elements of this Fuzzy Adaptive Neighborhood Search (FANS) algorithm is a fuzzy valuation, which is used to measure the degree to which the solutions that are considered at the decision stages accomplish a certain qualitative property. FANS is analyzed from two perspectives: first, it is shown how FANS may be adapted to behave like other traditional local search techniques by means of suitable definitions for the fuzzy valuation component. Second, comparisons are made to show the potential of the method as a general purpose optimization tool, when none or minimal knowledge of the problem being solved is available.Both aspects make FANS a valuable tool regarding further developments within the context of decision support systems involving heuristic algorithms.

Soft computing and cooperative strategies for optimization

In this paper we describe our work in the context of multi-thread, cooperative strategies to solve combinatorial optimization problems. Under our approach, we use soft computing ideas at two levels: to define the solvers or threads, and to define the rules governing the (centralized) coordination scheme. We show the results obtained regarding how the cooperative strategy behaves against a sequential and independent search scheme, and we analyze how the number of threads affects the search behavior.