A coevolutionary algorithm as a search strategy adaptation procedure in constrained optimization is discussed in the paper. The coevolutionary algorithm consists of the set of individual conventional genetic algorithms with different search strategies. Individual genetic algorithms compete and cooperate with each other. Competition is provided with resource re-allocation among algorithms and cooperation is provided with migration of the best individuals to all of the algorithms. At early works this method was applied for unconstrained optimization problems. The common result was that coevolutionary algorithm is more effective than average individual genetic algorithms. In this paper modification of competitive-cooperative coevolutionary algorithm for constrained optimization problems is considered. Results of test comparison of coevolutionary algorithm with conventional genetic algorithms demonstrate that coevolutionary algorithm is not less effective than the best for problem-in-hand individual conventional algorithm.
Abstract:The text classification problem for natural language call routing was considered in the paper. Seven different term weighting methods were applied. As dimensionality reduction methods, the feature selection based on self-adaptive GA is considered. k-NN, linear SVM and ANN were used as classification algorithms. The tasks of the research are the following: perform research of text classification for natural language call routing with different term weighting methods and classification algorithms and investigate the feature selection method based on self-adaptive GA. The numerical results showed that the most effective term weighting is TRR. The most effective classification algorithm is ANN. Feature selection with self-adaptive GA provides improvement of classification effectiveness and significant dimensionality reduction with all term weighting methods and with all classification algorithms. IntroductionNatural language call routing is an important problem in the design of modern automatic call services and the solving of this problem could lead to improvement of the call service [21]. Generally natural language call routing can be considered as two different problems. The first one is speech recognition of calls and the second one is topic categorization of users utterances for further routing. Topic categorization of users utterances can be also useful for multidomain spoken dialogue system design [12]. In this work we treat call routing as an example of a text classification application.In the vector space model [16] text classification is considered as a machine learning problem. The complexity of text categorization with a vector space model is compounded by the need to extract the numerical data from text information before applying machine learning algorithms. Therefore, text classification consists of two parts: text preprocessing and classification algorithm application using the obtained numerical data. Text preprocessing comprises three stages:-Textual feature extraction.-Term weighting -Dimensionality reduction. The first one is the textual feature extraction based on raw preprocessing of the documents. This process includes deleting punctuation, transforming capital letters to lowercase, and additional procedures such as stop-words filtering [4] and stemming [14]. Stop-words list contains pronouns, prepositions, articles and other words that usually have no importance for the classification. Using stemming it is possible to join different forms of the same word into one textual feature.The second stage is the numerical feature extraction based on term weighting. For term weighting we use "bag-of-words" model, in which the word order is ignored. There exist different unsupervised and supervised term weighting methods. The most well-known unsupervised term weighting method is TFIDF [15]. The following supervised term weighting methods are also considered in the paper:
A new method of Michigan and Pittsburgh approaches combination for fuzzy classifier design with evolutionary algorithms is presented. Fuzzy classifier design includes of four stages. The first stage is standard fuzzification. The second one is a special procedure of initial rules forming with a priori information from a learning sample. At the third stage Michigan method is applied and it provides fast search of fuzzy rules with the best grade of certainty values for different classes and smoothing of randomness at initial population forming. At the fourth stage Pittsburgh method provides rules subset search with the best performance and predefined number of the rules and doesn't require a lot of computational power. Besides, a self-tuning cooperative-competitive coevolutionary algorithm for strategy adaptation is applied at Michigan and Pittsburgh stages of the fuzzy classifier design. This algorithm automatically solves the problem of genetic algorithm parameters setting. Thereby the method allows getting a compact fuzzy rule set with appropriate classification performance and with high computation speed. Classification results for machine learning problems from UCI repository and comparison with different alternative classifiers are presented.
The quality of operation of neural networks in solving application problems is determined by the success of the stage of their training. The task of learning neural networks is a complex optimization task. Traditional learning algorithms have a number of disadvantages, such as «sticking» in local minimums and a low convergence rate. Modern approaches are based on solving the problems of adjusting the weights of neural networks using metaheuristic algorithms. Therefore, the problem of selecting the optimal set of values of algorithm parameters is important for solving application problems with symmetry properties. This paper studies the application of a new metaheuristic optimization algorithm for weights adjustment—the algorithm of the spiders-cycle, developed by the authors of this article. The approbation of the proposed approach is carried out to adjust the weights of recurrent neural networks used to solve the time series forecasting problem on the example of three different datasets. The results are compared with the results of neural networks trained by the algorithm of the reverse propagation of the error, as well as three other metaheuristic algorithms: particle swarm optimization, bats, and differential evolution. As performance criteria for the comparison of algorithms of global optimization, in this work, descriptive statistics for metrics of the estimation of quality of predictive models, as well as the number of calculations of the target function, are used. The values of the MSE and MAE metrics on the studied datasets were obtained by adjusting the weights of the neural networks using the cycling spider algorithm at 1.32, 25.48, 8.34 and 0.38, 2.18, 1.36, respectively. Compared to the inverse error propagation algorithm, the cycling spider algorithm reduced the value of the error metrics. According to the results of the study, it is concluded that the developed algorithm showed high results and, in the assessment of performance, was not inferior to the existing algorithm.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.