The identification of appropriate reaction models is very helpful for developing chemical vapor deposition (CVD) processes. In this study, we developed an automatic modeling system that analyzes experimental data on t...
详细信息
The identification of appropriate reaction models is very helpful for developing chemical vapor deposition (CVD) processes. In this study, we developed an automatic modeling system that analyzes experimental data on the cross- sectional shapes of films deposited on substrates with nanometer- or micrometer-sized trenches. The system then identifies a suitable reaction model to describe the film deposition. The inference engine used by the system to model the reaction mechanism was designed using real-coded genetic algorithms (RCGAs): a generation alternation model named “just generation gap” (JGG) and a real-coded crossover named “real-coded ensemble crossover” (REX). We studied the effect of REX+JGG on the system's performance, and found that the system with REX+JGG was the most accurate and reliable at model identification among the algorithms that we studied.
Because evolutionary algorithms (EAs) generally require many repeated evaluations of objective functions, it often takes considerable time to solve optimization problems. Parallel computation is one means to shorten t...
详细信息
Because evolutionary algorithms (EAs) generally require many repeated evaluations of objective functions, it often takes considerable time to solve optimization problems. Parallel computation is one means to shorten the required computation time. In earlier works, the authors proposed an EA suitable for coarse-grained parallel computers, a genetic local search with distance independent diversity control (GLSDC). Though GLSDC has been applied successfully to several practical problems, its parallel efficiency abruptly drops off as the number of CPUs for computation increases. To achieve a higher parallel efficiency, the authors now propose a new EA, an asynchronous GLSDC (AGLSDC), constructed by reworking the algorithm of GLSDC. This paper introduces the proposed method and reports verification of the method through numerical experiments on several benchmark problems and a practical problem.
The k-nearest neighbor (k-NN) algorithm is commonly used in applications of classifiers and data mining and (he related area due to its simplicity and effectiveness. In this study, all of features and optimal feature ...
详细信息
The k-nearest neighbor (k-NN) algorithm is commonly used in applications of classifiers and data mining and (he related area due to its simplicity and effectiveness. In this study, all of features and optimal feature subsets with three features are investigated. For classification, crisp k-NN, fuzzy k-NN, and weighting fuzzy k-NN classifiers are compared. For weighting of features, two types of coding including binary-codedgeneticalgorithms (BGA) and real-coded genetic algorithms (BGA) are evaluated. Experiments are conducted on the Wisconsin diagnosis breast cancer (WDBC) dataset and the Pima (PIMA) Indians diabetes dataset, and the classification accuracy, false negative, and computation time are reported in this paper.
暂无评论