APPLICATION OF DECISION TREE IN EVALUATING OF ENGINEERING QUALITY OF ROCK MASSES

• 1、School of Civil Engineering and Mechanics, Lanzhou University
• 2、School of Economy and Trade, Lanzhou University of Finance and Economics

Abstract：Rock mass classification is a typically nonlinear issue concerned with input and output, which is related to several uncertain factors. Distinct methods of rock mass classification refer to different indexes, which can have a great influence on the evaluation results. As a data mining classified arithmetic, decision tree can solve these problems effectively. Based on C4.5 algorithm and a selection of five attributes of the category attributes, a decision tree of rock mass classification is established，including six indexes：RQD(rock quality designation), Kv(integrity coefficient), Vp(longitudinal wave velocity), D(Joint spacing), Rc(uniaxial compressive strength)and ω(permeability coefficient of joints),which were represent Rock strength, Rock mass structure and Groundwater. The selection of the indexes based on analysis of some typical methods of rock mass classification. At last the decision tree is tested in example classifications of some rock masses, and the results fit well with the classification of Engineering Investigation，which indicates that this method has a highly practical application. At the same time the method offers a new thought to the research of rock mass classification.

Keywords： rock mass classification decision tree C4.5 algorithm engineering application