Rough K-means Clustering Algorithm And Its Application In Cultural Relics Health Evaluation

Cultural relics are important of human civilization,with historical,cultural and economic value.In the numerous cultural relics,open-air stone relics deposit in nature environment for a long time,facing a variety of health risks.The evaluation of cultural relics provides data support for the protection of cultural relics by means of prevention.It is a new idea and method in the protection of cultural relics by using the Internet of Things to collect natural environment data of cultural relics in real time,discovering the health status of the relics through data mining of the collected data.Cluster analysis is an important method in the field of data mining,and the data objects in the data set are clustered into different classes according to the rules.Rough Kmeans clustering algorithm can effectively deal with uncertain,incomplete and inaccurate data.But the rough K-means clustering algorithm exists defects,and these defects limit the use of rough K-means clustering algorithms.The main research work of this thesis:First,improved the rough K-means clustering algorithm.By using the ant colony algorithm to overcome the shortcomings of rough K-means clustering algorithm,in which the clustering result in initial selection of the initial point may leads to local optimal rather than global optimal.Based on previous research results,automatically calculating threshold and approximate area of weight,it is proposed an improved algorithm of the objective function with avoiding human implications of setting thresholds and weights of clustering results of previous research.The objective function considers the distance between the class and the distance within class.Besides,for the low clustering efficient of rough K-means clustering algorithm,dealing with incremental data in dynamic data set,it is proposed a method that silhouette coefficient handle with incremental data.And the stop principle of this method is converging the objective function.Second,the natural environment data of TangShun Ling open-air stone artifacts are analyzed,and the health evaluation indexes is established.An improved rough K-means clustering algorithm is used to cluster the natural environment data of Tang Shun Ling,and get stone health risk level according to the comprehensive index method.