Ensembles of classifiers is a way to improve the performance of the approach with single classifiers. The idea is to find and combine a set of classifiers that are responsible for smaller and theoretically easier part...
详细信息
ISBN:
(纸本)9781509035663
Ensembles of classifiers is a way to improve the performance of the approach with single classifiers. The idea is to find and combine a set of classifiers that are responsible for smaller and theoretically easier parts of a problem to solve, in other words, divide to conquer. Between the ensembles models, there is the clustering and selection in which the training data are clustering, and a classifier is built for each cluster found. An answer for an input data is given based on a distance to the available clusters that has an associated classifier. In this paper, the clustering and selection model is explored with the use of evolutionary algorithms to search clusters that optimize the ensemble's performance. Experiments are conducted with ten datasets and using recent advances in classification methods. The results achieved good and promising performances compared to classical clustering-and-selection model and other methods to build ensembles.
evolutionary algorithms (EAs), when used for global optimization, can be seen as unconstrained optimization techniques. Therefore, they require an additional mechanism to incorporate constraints of any kind (i.e., ine...
详细信息
ISBN:
(纸本)9781450343237
evolutionary algorithms (EAs), when used for global optimization, can be seen as unconstrained optimization techniques. Therefore, they require an additional mechanism to incorporate constraints of any kind (i.e., inequality, equality, linear, nonlinear) into their fitness function. Although the use of penalty functions (very popular with mathematical programming techniques) may seem an obvious choice, this sort of approach requires a careful fine tuning of the penalty factors to be used. Otherwise, an EA may be unable to reach the feasible region (if the penalty is too low) or may reach quickly the feasible region but being unable to locate solutions that lie in the boundary with the infeasible region (if the penalty is too severe). This has motivated the development of a number of approaches to incorporate constraints into the fitness function of an EA. This tutorial will cover the main proposals in current use, including novel approaches such as the use of tournament rules based on feasibility, multiobjective optimization concepts, hybrids with mathematical programming techniques (e.g., Lagrange multipliers), cultural algorithms, and artificial immune systems, among others. Other topics such as the importance of maintaining diversity, current benchmarks and the use of alternative search engines (e.g., particle swarm optimization, differential evolution, evolution strategies, etc.) will be also discussed (as time allows).
Online shopping has become an essential part of our life, which provides a suitable, cheap, and quick way for customers to enjoy a wide variety of products. However, due to the large number of online stores, a custome...
详细信息
ISBN:
(纸本)9781509035496
Online shopping has become an essential part of our life, which provides a suitable, cheap, and quick way for customers to enjoy a wide variety of products. However, due to the large number of online stores, a customer usually faces difficulties to review all available offers manually in order to find a favorite item. The Internet shopping optimization problem (ISOP) is a multiple-item multiple-shop optimization problem, which targets to minimize the total cost for a costumer to purchase a given set of products over all available offers. In this paper, the mathematical model of existing ISOP has been improved. In the improved model of ISOP different constraints and assumptions such as the maximum budget, discounts offered by internet shops have been taken into account. Several metaheuristic optimization methods such as the genetic algorithm are implemented. The obtained numerical results illustrate the effectiveness of the improved model and metaheuristics applied.
Finding the best possible sequence of control actions for a hybrid car in order to minimize fuel consumption is a well-studied problem. A standard method is Dynamic Programming (DP) that is generally considered to pro...
详细信息
ISBN:
(纸本)9783319312040
Finding the best possible sequence of control actions for a hybrid car in order to minimize fuel consumption is a well-studied problem. A standard method is Dynamic Programming (DP) that is generally considered to provide solutions close to the global optimum in relatively short time. To our knowledge evolutionary algorithms (EAs) have so far not been used for this setting, due to the success of DP. In this work we compare DP and EA for a well-studied example and find that for the basic scenario EA is indeed clearly outperformed by DP in terms of calculation time and quality of solutions. But, we also find that when going beyond the standard scenario towards more realistic (and complex) scenarios, EAs can actually deliver a performance en par or in some cases even exceeding DP, making them useful in a number of relevant application scenarios.
Genetic association is a challenging task for the identification and characterization of genes that increase the susceptibility to common complex multifactorial diseases. To fully execute genetic studies of complex di...
详细信息
Genetic association is a challenging task for the identification and characterization of genes that increase the susceptibility to common complex multifactorial diseases. To fully execute genetic studies of complex diseases, modern geneticists face the challenge of detecting interactions between loci. In this paper, two evolutionary methods were compared to detect associations of single nucleotide polymorphisms (SNPs): a genetic algorithm and Gauss particle swarm optimization. Genetic algorithm was developed with partial matched crossover operator and two different strategies for initialization: regular initialization and top-5 strategy initialization. In both methods for different SNP barcodes (SNP combinations with their corresponding genotypes) the difference between case and control data is computed systematically. The algorithms look for the best combination which is the barcode with maximum difference between the two groups. Analysis results support that the genetic algorithm with top-5 strategy for initialization provides higher frequency difference values than the Gauss particle swarm optimization. It is also proved that a genetic algorithm reduces a computational cost for obtaining higher frequency difference between the case and control group. (C) 2016 The Authors. Published by Elsevier B.V.
In the past decades, subsurface non-aqueous phase liquid (NAPL) contamination has been recognized as one of the most widespread and challenging environmental problems. Thus, researchers have focused their efforts on d...
详细信息
In the past decades, subsurface non-aqueous phase liquid (NAPL) contamination has been recognized as one of the most widespread and challenging environmental problems. Thus, researchers have focused their efforts on developing and testing the efficiency of remediation methodologies, able to address the unique nature of these contaminants. Recently, in-situ flooding techniques for the accelerated removal of NAPLs trapped in the subsurface have been proposed, where additives are injected together with water upgradient of the NAPL-contaminated area in order to alter the physio-chemical properties of the contaminants, such as interfacial tension, and enhance their solubilities. In this work, the efficiency of ethanol enhanced NAPL remediation is addressed. To this end, a non-linear, multi-objective optimization strategy is developed by combining a multiphase flow simulation model with evolutionary algorithms. Two conflicting optimization objectives are considered: minimizing operation cost and maximizing remediation efficiency, while preventing uncontrolled NAPL mobilization. More specifically, the first objective involves the operation cost of the procedure, which is directly proportional to the pumping rate, duration and ethanol volume used. The second represents the environmental considerations of the problem that, in this work, are described by the maximization of free product removal and the prevention of DNAPL vertical spreading. (C) 2016 The Authors. Published by Elsevier Ltd.
evolutionary algorithms have shown much success in solving real-world design problems, but they are considered computationally inefficient because they rely on many objective function evaluations instead of leveraging...
详细信息
ISBN:
(纸本)9780791850114
evolutionary algorithms have shown much success in solving real-world design problems, but they are considered computationally inefficient because they rely on many objective function evaluations instead of leveraging domain knowledge to guide the optimization. An evolutionary algorithm's performance can be improved by utilizing operators called domain specific heuristics that incorporate domain knowledge, but existing knowledge-intensive algorithms utilize one or two domain specific heuristics, which limits the amount of incorporated knowledge or treats all knowledge as equally effective. We propose a hyperheuristic approach that efficiently utilizes multiple domain-specific heuristics that incorporate knowledge from different sources by allocating computational resources to the effective ones. Furthermore, a hyperheuristic allows the simultaneous use of conventional evolutionary operators that assist in escaping local optima. This paper empirically demonstrates the efficacy of the proposed hyperheuristic approach on a multi-objective design problem for an Earth observation satellite system. Results show that the hyperheuristic approach significantly improves the search performance compared to an evolutionary algorithm that does not use any domain knowledge.
This paper analyzes the influence of specific genetic operations on the results obtained when applying the Unplugged evolutionary Algorithm in an artistic creation process. Throughout the methodology developed in prev...
详细信息
This paper presents an FPGA based synthesizable offline UAV local path planner implementation using evolutionary algorithms for 3D unknown environments. A Genetic Algorithm is selected as the path planning algorithm a...
详细信息
Bin packing problems are a class of optimization problems that have numerous applications in the industrial world, ranging from efficient cutting of material to packing various items in a larger container. We consider...
详细信息
ISBN:
(纸本)9780889869844
Bin packing problems are a class of optimization problems that have numerous applications in the industrial world, ranging from efficient cutting of material to packing various items in a larger container. We consider here only rectangular items cut off an infinite strip of material as well as off larger sheets of fixed dimensions. This problem has been around for many years and a great number of publications can be found on the subject. Nevertheless, it is often difficult to reconcile a theoretical paper and practical application of it. The present work aims to create simple but, at the same time, fast and efficient algorithms, which would allow one to write high-speed and capable software that can be used in a real-time application.
暂无评论