Bayes Discriminant Analysis Model for Water Inrush Risk from Coal Seam Floor and its Application

Based on the basic idea of Bayes discriminant analysis method, we established Bayes discriminant analysis model for evaluating water inrush risk from coal seam floor. Five indicators such as aquifer property, water head, water-resisting layer thickness, fault permeability and fractures development level were used as discriminant factors in the model. Fourteen groups of measured data on coal seam from diffenent coal mines were regarded as training samples, then the Bayes discriminant model was set up. Fourteen groups of experimental data of coal seam were discriminated by back substitution discriminant method and then the model was used in practical engineering. The results show that misjudgment rate of the Bayes discriminant analysis model is low and it can effectively distinguish the coal floor water inrush risk level. So the model can be better applied to practical engineering.