A main challenge in evolutionary algorithms (EAs) is determining a termination condition ensuring stabilization close to the optimum in real-world applications. Although for known test functions distribution-based qua...
详细信息
Successful and efficient use of evolutionary algorithms (EAs) depends on the choice of the genotype, the problem representation (mapping from genotype to phenotype) and on the choice of search operators that are appli...
详细信息
ISBN:
(纸本)9781450306904
Successful and efficient use of evolutionary algorithms (EAs) depends on the choice of the genotype, the problem representation (mapping from genotype to phenotype) and on the choice of search operators that are applied to the genotypes. These choices cannot be made independently of each other. The question whether a certain representation leads to better performing EAs than an alternative representation can only be answered when the operators applied are taken into consideration. The reverse is also true: deciding between alternative operators is only meaningful for a given representation. In EA practice one can distinguish two complementary approaches. The first approach uses indirect representations where a solution is encoded in a standard data structure, such as strings, vectors, or discrete permutations, and standard off-the-shelf search operators are applied to these genotypes. To evaluate the solution, the genotype needs to be mapped to the phenotype space. The proper choice of this genotype-phenotype mapping is important for the performance of the EA search process. The second approach, the direct representation, encodes solutions to the problem in its most 'natural' space and designs search operators to operate on this representation. Research in the last few years has identified a number of key concepts to analyse the influence of representation-operator combinations on EA performance. These concepts are*locality and*redundancy. Locality is a result of the interplay between the search operator and the genotype-phenotype mapping. Representations are redundant if the number of phenotypes exceeds the number of possible genotypes. Furthermore, redundant representations can lead to biased encodings if some phenotypes are on average represented by a larger number of genotypes. Finally, a bias need not be the result of the representation but can also be caused by the search operator. The tutorial gives a brief overview about existing guidelines for representatio
this paper introduces the compressed air supply network based on the theory of evolution of the optimization scheduling control system. According to enterprise widespread surplus compressed air compressor, unloading o...
详细信息
The use of niching methods for solving real world optimization problems is limited by the difficulty to obtain a proper setting of the speciation parameters without any a priori information about the fitness landscape...
详细信息
ISBN:
(纸本)9781450305570
The use of niching methods for solving real world optimization problems is limited by the difficulty to obtain a proper setting of the speciation parameters without any a priori information about the fitness landscape. To avoid such a difficulty, we propose a novel method, called Adaptive Species Discovery, that removes the basic assumption of perfect discrimination among peaks underlying Fitness Sharing and, consequently, allows to overcome the drawbacks of the most performing sharing-based methods. This is achieved through an explicit mechanism able to discover the species in the population during the evolution. The method does not require any a priori knowledge, in that it makes no assumption about the location and the shape of the peaks, while it exploits information about the ruggedness of the fitness landscape, dynamically acquired at each generation. The proposed method has been evaluated on a set of standard functions largely adopted in the literature to assess the performance of niching methods. The experimental results show that our method has a better ability to discover and maintain all the peaks with respect to other methods proposed so far. Copyright 2011 ACM.
Memetic algorithms are evolutionary metaheuristics intrinsically concerned with the exploiting and incorporation of all available knowledge about the problem under study. In this dissertation, we present a knowledge-b...
详细信息
Memetic algorithms are evolutionary metaheuristics intrinsically concerned with the exploiting and incorporation of all available knowledge about the problem under study. In this dissertation, we present a knowledge-based memetic algorithm to tackle the threedimensional protein structure prediction problem without the explicit use of template experimentally determined structures. The algorithm was divided into two main steps of processing: (i) sampling and initialization of the algorithm solutions; and (ii) optimization of the structural models from the previous stage. The first step aims to generate and classify several structural models for a determined target protein, by the use of the strategy Angle Probability List, aiming the definition of different structural groups and the creation of better structures to initialize the initial individuals of the memetic algorithm. The Angle Probability List takes advantage of structural knowledge stored in the Protein Data Bank in order to reduce the complexity of the conformational search space. The second step of the method consists in the optimization process of the structures generated in the first stage, through the applying of the proposed memetic algorithm, which uses a tree-structured population, where each node can be seen as an independent subpopulation that interacts with others, over global search operations, aiming at information sharing, population diversity, and better exploration of the multimodal search space of the problem The method also encompasses ad-hoc global search operators, whose objective is to increase the exploration capacity of the method turning to the characteristics of the protein structure prediction problem, combined with the Artificial Bee Colony algorithm to be used as a local search technique applied to each node of the tree. The proposed algorithm was tested on a set of 24 amino acid sequences, as well as compared with two reference methods in the protein structure prediction area, Ros
Game playing is a relatively interesting task in the field of artificial intelligence in these days. The master thesis deals with general artificial intelligence which is capable of playing selected simple games based...
详细信息
Game playing is a relatively interesting task in the field of artificial intelligence in these days. The master thesis deals with general artificial intelligence which is capable of playing selected simple games based on information that is also avai- lable to the human player. Our selected games are 2048, Mario, racing simulator TORCS and Alhambra. All the information acquired by artificial intelligence is provided by games through an interface, therefore none of the models uses visual input. We use evolutionary approaches such as evolutionary algorithms, evolutio- nary strategy CMA and differential evolution applied to different types of neural networks. We are also dealing with deep reinforcement learning. We test these approaches and compare their results. 1
This paper investigates the effectiveness and efficiency of two competitive (predator-prey) evolutionaryprocedures for training multi-layer perceptron classifiers: Co-Adaptive Neural Network Training, and a modifiedve...
详细信息
This paper investigates the effectiveness and efficiency of two competitive (predator-prey) evolutionaryprocedures for training multi-layer perceptron classifiers: Co-Adaptive Neural Network Training, and a modifiedversion of Co-evolutionary Neural Network Training. The study focused on how the performance of the two procedures varies as the size of the training set increases, and their ability to redress class imbalance problems of increasing severity. Compared to the customary backpropagation algorithm and a standard evolutionary algorithm, the two competitive procedures excelled in terms of quality of the solutions and execution speed. Co-Adaptive Neural Network Training excelled on class imbalance problems, and on classification problems of moderately large training sets. Co-evolutionary Neural Network Training performed best on the largest data sets. The size of the training set was the most problematic issue for the backpropagation algorithm and the standard evolutionary algorithm, respectively in terms of accuracy of the solutions and execution speed. Backpropagation and the evolutionary algorithm were also not competitive on the class imbalance problems, where data oversampling could only partially remedy their shortcomings.
Wheel-legged hybrid robots are versatile machines that can employ several locomotion modes;however, automatically choosing the right locomotion mode is still an open problem in robotics. We here propose that the robot...
详细信息
Thin-film solar cells enable a strong reduction of the amount of silicon needed to produce photovoltaic panels but their efficiency lowers. Placing metallic or dielectric nanoparticles over the silicon substrate incre...
详细信息
Thin-film solar cells enable a strong reduction of the amount of silicon needed to produce photovoltaic panels but their efficiency lowers. Placing metallic or dielectric nanoparticles over the silicon substrate increases the light trapping into the panel thanks to the plasmonic scattering from nanoparticles at the surface of the cell. The goal of this paper is to optimize the geometry of a thin-film solar cell with silver and silica nanoparticles in order to improve its efficiency, taking into account the amount of silver. An efficient evolutionary algorithm is applied to perform the optimization with a reduced computing time.
In spite of the success of evolutionary algorithms for dealing with multi-objective optimization problems (the so-called multi-objective evolutionary algorithms (MOEAs)), their main drawback is the fine-tuning of thei...
详细信息
暂无评论