Research on Evaluation Method of Forest Land Based on Decision Tree and Bootstrap Regression

The reasonable evaluation of forestry land is the basis of marketization of the forestry assets. Based on the evaluation data for 325 forestry lands from seven forest farms of Ning´an city, which is intended to transfer in the country, this paper takes Larch, mongolian scots pine and larch-mongolian mixed forests as instances to establish CHAID decision tree model in order to classify and analyze forestry assets and find main qualitative influencing factors. On this basis, Bootstrap regressive model is applied to predict and evaluate forestry assets according to quantitative influencing factors. The practicability of method has been verified by calculations. The purpose of this paper is to put forward a reasonable evaluation method of forestry land based on data mining, which would perfect the evaluation system of forestry land and provide supporting data for marketization of forestry assets.