Nearest prototype classification of noisy data

Nearest prototype approaches offer a common way to design classifiers. However, when data is noisy, the success of this sort of classifiers depends on some parameters that the designer needs to tune, as the number of prototypes. In this work, we have made a study of the ENPC technique, based on the nearest prototype approach, in noisy datasets. Previous experimentation of this algorithm had shown that it does not require any parameter tuning to obtain good solutions in problems where class limits are well defined, and data is not noisy. In this work, we show that the algorithm is able to obtain solutions with high classification success even when data is noisy. A comparison with optimal (hand made) solutions and other different classification algorithms demonstrates the good performance of the ENPC algorithm in accuracy and number of prototypes as the noise level increases. We have performed experiments in four different datasets, each of them with different characteristics.     

Hybrid evolutionary method for obstacle location-allocation

In this paper we propose a hybrid evolutionary method for Obstacle Location-allocation problem. This problem can be described as a tri-level mixed integer programming problem. Since this problem is very complex and with many local solutions, no direct method is effective to solve it. Heuristic methods were proposed to it, but optimality is not guaranteed yet. Our hybrid evolutionary method adopts the main structure of Genetic Algorithms (GA) absorbing ideas from Evolutionary Strategy (ES) and combines with some traditional optimization techniques. In this way we can pursue global optimization maintaining a good efficiency of our method. A case study shows the effectiveness of this method.     

Local feature weighting in nearest prototype classification.

The distance metric is the corner stone of nearest neighbor (NN)-based methods, and therefore, of nearest prototype (NP) algorithms. That is because they classify depending on the similarity of the data. When the data is characterized by a set of features which may contribute to the classification task in different levels, feature weighting or selection is required, sometimes in a local sense. However, local weighting is typically restricted to NN approaches. In this paper, we introduce local feature weighting (LFW) in NP classification. LFW provides each prototype its own weight vector, opposite to typical global weighting methods found in the NP literature, where all the prototypes share the same one. Providing each prototype its own weight vector has a novel effect in the borders of the Voronoi regions generated: They become nonlinear. We have integrated LFW with a previously developed evolutionary nearest prototype classifier (ENPC). The experiments performed both in artificial and real data sets demonstrate that the resulting algorithm that we call LFW in nearest prototype classification (LFW-NPC) avoids overfitting on training data in domains where the features may have different contribution to the classification task in different areas of the feature space. This generalization capability is also reflected in automatically obtaining an accurate and reduced set of prototypes.     

Evolving gaits for hexapod robots using cyclic genetic algorithms

A major facet of multi-legged robot control is locomotion. Each leg must move in such a manner that it efficiently produces thrust and provides maximum support. The motion of all the legs must be coordinated so that they are working together to provide constant stability while propelling the robot forward. In this paper, we discuss the use of a cyclic genetic algorithm (CGA) to evolve control programs that produce gaits for actual hexapod robots. Tests done in simulation and verified on the actual robot show that the CGA successfully produces gaits for both fully capable and disabled robots.     

Estimating software testing complexity

ContextComplexity measures provide us some information about software artifacts. A measure of the difficulty of testing a piece of code could be very useful to take control about the test phase.ObjectiveThe aim in this paper is the definition of a new measure of the difficulty for a computer to generate test cases, we call it Branch Coverage Expectation (BCE). We also analyze the most common complexity measures and the most important features of a program. With this analysis we are trying to discover whether there exists a relationship between them and the code coverage of an automatically generated test suite.MethodThe definition of this measure is based on a Markov model of the program. This model is used not only to compute the BCE, but also to provide an estimation of the number of test cases needed to reach a given coverage level in the program. In order to check our proposal, we perform a theoretical validation and we carry out an empirical validation study using 2600 test programs.ResultsThe results show that the previously existing measures are not so useful to estimate the difficulty of testing a program, because they are not highly correlated with the code coverage. Our proposed measure is much more correlated with the code coverage than the existing complexity measures.ConclusionThe high correlation of our measure with the code coverage suggests that the BCE measure is a very promising way of measuring the difficulty to automatically test a program. Our proposed measure is useful for predicting the behavior of an automatic test case generator.     

Choosing selection pressure for wide-gap problems

To exploit an evolutionary algorithm’s performance to the full extent, the selection scheme should be chosen carefully. Empirically, it is commonly acknowledged that low selection pressure can prevent an evolutionary algorithm from premature convergence, and is thereby more suitable for wide-gap problems. However, there are few theoretical time complexity studies that actually give the conditions under which a high or a low selection pressure is better. In this paper, we provide a rigorous time complexity analysis showing that low selection pressure is better for the wide-gap problems with two optima.     

Computational modeling of city formation

Although a burst of recent research in economics has examined how industries form, a majority of it considers highly simplified models. In this paper, we use computational modeling techniques to expand from traditional, simple, analytically tractable economic models to more complex two dimensional landscapes. Using the basic theories developed in earlier research, we examine what factors cause cities to emerge, including: transportation costs, the percentage of workers in a population, and the elasticity of substitution. These three factors should cause workers and firms to agglomerate, causing cities to emerge out of a scattered population.     

Observations in using parallel and sequential evolutionary algorithms for automatic software testing

In this paper we analyze the application of parallel and sequential evolutionary algorithms (EAs) to the automatic test data generation problem. The problem consists of automatically creating a set of input data to test a program. This is a fundamental step in software development and a time consuming task in existing software companies. Canonical sequential EAs have been used in the past for this task. We explore here the use of parallel EAs. Evidence of greater efficiency, larger diversity maintenance, additional availability of memory/CPU, and multi-solution capabilities of the parallel approach, reinforce the importance of the advances in research with these algorithms. We describe in this work how canonical genetic algorithms (GAs) and evolutionary strategies (ESs) can help in software testing, and what the advantages are (if any) of using decentralized populations in these techniques. In addition, we study the influence of some parameters of the proposed test data generator in the results. For the experiments we use a large benchmark composed of twelve programs that includes fundamental algorithms in computer science.     

An evolutionary approach to deception in multi-agent systems

Understanding issues of trust and deception are key to designing robust, reliable multi-agent systems. This paper builds on previous work which examined the use of auctions as a model for exploring the concept of deception in such systems. We have previously described two forms of deceptive behaviour which can occur in a simulated repeated English auction. The first of these types of deception involves sniping or late bidding, which not only allows an agent to conceal its true valuation for an item, but also potentially allows it to win an item for which it may not possess the highest valuation. The second deceptive strategy involves the placing of false bids which are designed to reduce an opponent’s potential profit. In this work we examine the potential shortcomings of those two strategies and investigate whether or not their individual strengths can be combined to produce a successful hybrid deceptive strategy.     

Self-emergence of a common lexicon by evolution in teams of autonomous agents

In multi-agent systems, the study of language and communication is an active field of research. In this paper we present the application of evolutionary strategies to the self-emergence of a common lexicon in a population of agents. By modeling the vocabulary or lexicon of each agent as an association matrix or look-up table that maps the meanings (i.e. the objects encountered by the agents or the states of the environment itself) into symbols or signals we check whether it is possible for the population to converge in an autonomous, decentralized way to a common lexicon, so that the communication efficiency of the entire population is optimal. We have conducted several experiments aimed at testing whether it is possible to converge with evolutionary strategies to an optimal Saussurean communication system. We have organized our experiments alongside two main lines: first, we have investigated the effect of the population size on the convergence results. Second, and foremost, we have also investigated the effect of the lexicon size on the convergence results. To analyze the convergence of the population of agents we have defined the population's consensus when all the agents (i.e. 100% of the population) share the same association matrix or lexicon. As a general conclusion we have shown that evolutionary strategies are powerful enough optimizers to guarantee the convergence to lexicon consensus in a population of autonomous agents.     

Where is evolutionary computation going? A temporal analysis of the EC community

Studying an evolving complex system and drawing some conclusions from it is an integral part of nature-inspired computing; being a part of that complex system, some insight can also be gained from our knowledge of it. In this paper we study the evolution of the evolutionary computation co-authorship network using social network analysis tools, with the aim of extracting some conclusions on its mechanisms. In order to do this, we first examine the evolution of macroscopic properties of the EC co-authorship graph, and then we look at its community structure and its corresponding change along time. The EC network is shown to be in a strongly expansive phase, exhibiting distinctive growth patterns, both at the macroscopic and the mesoscopic level.

Hybrid evolutionary method for capacitated location-allocation problem

Location-allocation model is widely applied for facility location design in practice. In this paper, we discuss an extension of location-allocation model which has capacity constraints and propose a hybrid evolutionary method to solve it which absorbs ideas from both genetic algorithms (GAs) and evolutionary strategy (ES) as well as combined with efficient traditional optimization techniques. It is shown that the proposed method is effective in finding global or near global solutions by numerical simulations.     

A cascade evolutionary algorithm for the bodyguard allocation problem

The bodyguard allocation problem (BAP) is an optimization problem that illustrates the behavior of processes with contradictory individual goals in some distributed systems. The objective function of this problem is the maximization of a parameter called the social welfare. Although the main method proposed to solve this problem, known as CBAP, is simple and time efficient, it lacks the ability to generate a diverse set of solutions, which is one of the most important feature to improve the chances to reach the global optimum. To overcome this drawback, we address the BAP with an evolutionary algorithm, the EBAP. Later, we take advantage of the best properties of both algorithms, EBAP and CBAP, to generate a two-stage cascade evolutionary algorithm called FFC-BAP. Extensive experimental results show that the algorithm FFC-BAP outperforms both the EBAP and the CBAP, in terms of quality of solutions.     

PITAGORAS-PSP: Including domain knowledge in a multi-objective approach for protein structure prediction

Whenever evolutionary algorithms are used to solve certain classes of problems such as those that present a huge search space, the incorporation of problem-specific knowledge is required to achieve adequate levels of performance. In this paper, we propose a multi-objective optimization-based procedure that includes such a domain-specific knowledge to cope with a difficult problem, the protein structure prediction (PSP). This problem is considered to be an open problem as there is no recognized “best” procedure to find solutions. It presents a vast search space and the analysis of each protein conformation requires significant amount of computing time. In our procedure, we provide a reduction of the search space by using the dependent rotamer library and include new heuristics to improve a multi-objective approach to PSP based on the PAES algorithm. As it is shown in the paper, by using benchmark proteins from the CASP8 set, this hybrid PSP procedure provides competitive results when it is compared with some of the better proposals appeared up to now.     

演化硬件技术及其在硬件设计中的应用

本文给出了基于演化算法和可编程逻辑器件的演化硬件的实现原理，并结合一个应用 实例加以说明，分析了该方法与传统的硬件设计方法相比的优越性，最后扼要介绍了演化硬件的研究方法以及它的应用现状和发展方向。     

An improved evolutionary algorithm for handling many-objective optimization problems

It has been shown that the multi-objective evolutionary algorithms (MOEAs) act poorly in solving many-objective optimization problems which include more than three objectives. The research emphasis, in recent years, has been put into improving the MOEAs to enable them to solve many-objective optimization problems efficiently. In this paper, we propose a new composite fitness evaluation function, in a novel way, to select quality solutions from the objective space of a many-objective optimization problem. Using this composite function, we develop a new algorithm on a well-known NSGA-II and call it FR-NSGA-II, a fast reference point based NSGA-II. The algorithm is evaluated for producing quality solutions measured in terms of proximity, diversity and computational time. The working logic of the algorithm is explained using a bi-objective linear programming problem. Then we test the algorithm using experiments with benchmark problems from DTLZ family. We also compare FR-NSGA-II with four competitive algorithms from the extant literature to show that FR-NSGA-II will produce quality solutions even if the number of objectives is as high as 20.     

Evolutionary approach for integration of multiple pronunciation patterns for enhancement of dysarthric speech recognition

Dysarthria is a motor speech disorder caused by neurological injury of the motor component of the motor-speech system. Because it affects respiration, phonation, and articulation, it leads to different types of impairments in intelligibility, audibility, and efficiency of vocal communication. Speech Assistive Technology (SAT) has been developed with different approaches for dysarthric speech and in this paper we focus on the approach that is based on modeling of pronunciation patterns. We present an approach that integrates multiple pronunciation patterns for enhancement of dysarthric speech recognition. This integration is performed by weighting the responses of an Automatic Speech Recognition (ASR) system when different language model restrictions are set. The weight for each response is estimated by a Genetic Algorithm (GA) that also optimizes the structure of the implementation technique (Metamodels) which is based on discrete Hidden Markov Models (HMMs). The GA makes use of dynamic uniform mutation/crossover to further diversify the candidate sets of weights and structures to improve the performance of the Metamodels. To test the approach with a larger vocabulary than in previous works, we orthographically and phonetically labeled extended acoustic resources from the Nemours database of dysarthric speech. ASR tests on these resources with the proposed approach showed recognition accuracies over those obtained with standard Metamodels and a well used speaker adaptation technique. These results were statistically significant.     

A new approach to estimating the expected first hitting time of evolutionary algorithms

Evolutionary algorithms (EA) have been shown to be very effective in solving practical problems, yet many important theoretical issues of them are not clear. The expected first hitting time is one of the most important theoretical issues of evolutionary algorithms, since it implies the average computational time complexity. In this paper, we establish a bridge between the expected first hitting time and another important theoretical issue, i.e., convergence rate. Through this bridge, we propose a new general approach to estimating the expected first hitting time. Using this approach, we analyze EAs with different configurations, including three mutation operators, with/without population, a recombination operator and a time variant mutation operator, on a hard problem. The results show that the proposed approach is helpful for analyzing a broad range of evolutionary algorithms. Moreover, we give an explanation of what makes a problem hard to EAs, and based on the recognition, we prove the hardness of a general problem.     

Improving nearest neighbor classification using Ensembles of Evolutionary Generated Prototype Subsets

One of the most accurate types of prototype selection algorithms, preprocessing techniques that select a subset of instances from the data before applying nearest neighbor classification to it, are evolutionary approaches. These algorithms result in very high accuracy and reduction rates, but unfortunately come at a substantial computational cost. In this paper, we introduce a framework that allows to efficiently use the intermediary results of the prototype selection algorithms to further increase their accuracy performance. Instead of only using the fittest prototype subset generated by the evolutionary algorithm, we use multiple prototype subsets in an ensemble setting. Secondly, in order to classify a test instance, we only use prototype subsets that accurately classify training instances in the neighborhood of that test instance. In an experimental evaluation, we apply our new framework to four state-of-the-art prototype selection algorithms and show that, by using our framework, more accurate results are obtained after less evaluations of the prototype selection method. We also present a case study with a prototype generation algorithm, showing that our framework is easily extended to other preprocessing paradigms as well.     

Periodical switching between related goals for improving evolvability to a fixed goal in multi-objective problems

Evolutionary computation plays a principal role in intelligent design automation. Evolutionary approaches have discovered novel and patentable designs. Nonetheless, evolutionary techniques may lead to designs that lack robustness. This critical issue is strongly connected to the concept of evolvability. In nature, highly evolvable species tend to be found in rapidly changing environments. Such species can be considered robust against environmental changes. Consequently, to create robust engineering designs it could be beneficial to use variable, rather than fixed, fitness criteria. In this paper, we study the performance of an evolutionary programming algorithm with periodical switching between goals, which are selected randomly from a set of related goals. It is shown by a dual-objective filter optimization example that altering goals may improve evolvability to a fixed goal by enhancing the dynamics of solution population, and guiding the search to areas where improved solutions are likely to be found. Our reference algorithm with a single goal is able to find solutions with competitive fitness, but these solutions are results of premature convergence, because they are poorly evolvable. By using the same algorithm with switching goals, we can extend the productive search length considerably; both the fitness and robustness of such designs are improved.