Miyun Forest Data Mining And Analysis Of Site Characteristics Of Main Tree Species

The implicit and useful information extracting from the mass of forest inventory data will enable the investigation of forest resources become more meaningful. It will provide technical assistant for forestry construction, operation and decision making with scientific data and systematically research. In this paper, the research is based on the data of Miyun forest resources investigation in 2009. Using data mining in three aspects:the overall Miyun forest survey data, dominant tree species data, and land types data. Clustering analysis and association analysis module in SPSS Clementine 12.0 were used for data mining. Clustering analysis was realized by Kohonen neural network clustering algorithm, K-Means clustering algorithm and TwoStep clustering algorithm. Correlation analysis is using Apriori correlation algorithm and GRI algorithm.Miyun forest subcompartment overall analysis to conclude information of the growth situation, the best site condition, environment characteristics and existing forest category, land category etc. Miyun Association of site condition. Study on Pinus tabulaeformis, Platycladus orientalis, Robinia pseudoacacia and Quercus palustris were concluded that the growth status of each tree species of the best forest subcompartment environment and site conditions, in order to draw the most suitable conditions for this kind of tree. Correlation analysis educed the correlation characteristics of forest subcompartment site conditions where this kind of tree growing. We made cluster analysis on coniferous forest and broadleaf forest and have obtained the most suitable forest subcompartment for them in Miyun.