학술논문

Application of improved CART algorithm in prediction of water inrush from coal seam floor
Document Type
article
Source
Gong-kuang zidonghua, Vol 40, Iss 12, Pp 52-56 (2014)
Subject
coal seam floor
water inrush prediction
cart algorithm
decision tree model
optimal threshold
Mining engineering. Metallurgy
TN1-997
Language
Chinese
ISSN
1671-251X
1671-251x
Abstract
In view of disadvantages of long time-consuming and low prediction accuracy existed in prediction model of water inrush from coal seam floor based on traditional CART algorithm, a decision tree model based on improved CART algorithm was introduced, and the decision tree model was used to establish prediction model of water inrush from coal seam floor. The experiment results show that running time of the prediction model of water inrush from coal seam floor based on improved CART algorithm is reduced from 1.0411 s to 0.6125 s, and prediction accuracy rate is improved from 88.78% to 95.54%.