Abstract

Extracting intuitive and useful information from the high-dimensional, fuzzy and complex operational simulation training data is in urgent need. In this paper, the operational simulation training data refers to the quantitative, numeric data that is usually used as the simulation results. The traditional statistical analysis methods have some limitations in clustering, visualizing and evaluating the operational simulation training data. We propose a new fuzzy clustering based method for evaluating the training effect of operational simulation training. Data clustering is an important technique for statistical data analysis and it enables the identification of a finite set of categories or clusters to describe the high-dimensional operational simulation training data. The classical partitional clustering algorithm, such as fuzzy c-means, is sensitive to the selection of the initial centers and may converge to a local minimum of the criterion function value if the initial centers are not properly chosen. Furthermore, the classical partitional clustering algorithm also has the problem of being unable to identify cluster centers that are uniformly distributed in a manifold. To solve these problems, we propose a geodesic distance and centrality based fuzzy c-Medoid clustering algorithm. Then we propose a fuzzy transitive closure based comparison algorithm to grade the cluster centers to get the qualitative synthetical grades. The method is demonstrated to be effective by an example of a virtual simulation training system of earthquake rescue.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call