Applied Mechanics and Materials Vols. 687-691

Abstract: Structural equation model is the only statistical methods that can simultaneously do comprehensive examination about the relationships between the multi-dimensional variables. It has broad application prospects in the field of applied linguistics. In this paper, combined with the applied linguistics article that choose structural equation modeling as the technique on academic journals over the past decades, then discussing the differences in domestic and foreign researchers from both content and form, after comparison and analysis, putting forward the suggestions for the application of SEM in our domestic applied linguistics.

Abstract: Today's very popular search engine technology, which is conducive to further analysis and full-text retrieval technology for Chinese word segmentation technology, and Chinese word segmentation is an important technology of Chinese information, the quality of Chinese word segmentation will have a direct impact on the efficiency of Chinese information. Therefore, the related concepts of the Chinese algorithm are discussed in this paper, some specific algorithm for Chinese, like algorithms based on rules and dictionary, statistical algorithms based on large-scale corpus, unity algorithm of statistics and the rule, artificial intelligence word segmentation algorithms and so on, and finally it describes the evaluated basis and difficulty of Chinese word segmentation algorithm.

Abstract: Under the influence of network information resources in exponentially growing, we are entering the information society, and all aspects of our lives has permeated with information technology. But the Chinese word segmentation technology for processing Chinese information becomes more and more important. Therefore, this paper sets out a series of Chinese word segmentation techniques, which mainly consists of Chinese word segmentation technology based on statistic, Chinese word segmentation techniques based on dictionary and hybrid techniques of Chinese word segmentation and segmentation technology based on knowledge and understanding.

Abstract: The mathematic element and process step of regress analysis, regress equation signification test and regress parameter signification test in uniform design experiment are analyzed. The regress analysis, regress equation signification test and regress parameter signification test of one uniform design experiment’s data are realized using MATLAB. The regress equation is gained, and significant factors are found. The computed conclusion is analyzed.

Abstract: Aiming at the bad performance when achieve rich colors of fabric with very limited yarns in the traditional woven industry, the paper comes up with a solution of selecting yarn from a set of yarns based on SAGA(simulated annealing genetic algorithm). In order to reduce the computational complexity, original image is compressed based on clustering algorithm. And the original yarns is divided into four regions based on color separation algorithm to narrow the feasible area. The result of experiments show that image compression and yarns division can greatly improve the speed of SAGA, and SAGA can effectively converges to global optimal solution.

Abstract: Aiming at the optimal decision problem of multi factors, model is proposed of establishing BN with cause-and-effect relationships, constructing evaluation function with multi factors by decision objectives, finally comparing and selecting the optimal solution of evaluation function from all possible decisions. The model inherits the ability of numerical reasoning and can analyze multi factors. Example illuminates procedures of the model in the paper.

Abstract: Invasive weed optimization algorithm is a new swarm intelligence algorithm recently. The algorithm has better robustness and adaptation, which is a very good intelligent optimization tools; but it is easy to fall into local optimization, and having the low speed of convergence, and it can not acquire exactly. Aiming at the shortcomings of the algorithm, taking advantage of pattern search excellent local search ability, this paper presents a novel hybrid optimization algorithm of pattern search algorithm and IWO optimization. The Simulation results of three standard benchmark functions show that the improved algorithm can greatly improve the convergence precision and convergence speed, and can effectively discourage the premature convergence.

Abstract: This paper presents three models as a potential decision making method for selecting the best baseball, field hockey, and women’s basketball NCAA Division I coaches. Five indicators, synthesized coaching efficiency, winning percentage, consecutive championship, achievement index and gender, are introduced to give a comprehensive evaluation of coaching ability. The preliminary served as a filter model to screen out less capable coaches and a robust ranking within top 10 is achieved. The Data Envelopment Analysis (DEA) model takes the time line horizon into consideration, and helps find less efficient coaches. By comparing the first two models, we obtain a reasonable assessment of coaches from different time period. Finally, by applying the Analytical Hierarchy Process (AHP), minor changes in judgment matrices can be made to adjust the ratio of male to female in the top 5 coaches.

Abstract: This paper develops a model validation method in the case of the full scale tests for the system model are infeasible. The Bayesian network with uncertain conditional probability parameters is used to represent the relations between the large computational model and its smaller modules. The interval probability theory is adopted to extrapolate the posterior probability of the interested variable in the uncertain Bayesian network. An interval valued Bayes factor is obtained to be the metric for model validation.

Abstract: Evolutionary algorithm is a random search method. In recent decades, various evolutionary algorithms emerge in endlessly by imitating population activity or the law of nature, so that evolutionary computation has been developed rapidly. The study of mathematical model about evolutionary algorithm is very lack. Only a small amount of research has been done. Through the study of the matrix model of evolutionary computation, this paper discusses the evolutionary process based on the matrix model in detail, and the working principle of matrix model and physical significance. The improvement suggestions of evolutionary algorithm have been given from the angle of the matrix model.