Skip to Main Content
Skip Nav Destination
ASME Press Select Proceedings
Geological Engineering: Proceedings of the 1st International Conference (ICGE 2007)
Baosong Ma
Baosong Ma
Search for other works by this author on:
No. of Pages:
ASME Press
Publication date:

Aiming at the complexity and spatial variability of the water inrush from coal floor, the paper proposed a method of appraising the risk of water inrush based on coupled SVM with GIS. Firstly, using the new machine learning tool-support vector machine, the paper presents a new method of forecast of water inrush from coal floor based on least square support vector machine and constructs the prediction model. Overcoming the extra-learning problem of ANN, the complicated nonlinear relationship between the water inrush risk and its affected actors is presented as well. The SVM following the principle of structure risk minimization, the paper also analyzed how kernel parameter σ and penal factor C affect the forecast accuracy. Then the SVM-GIS coupled appraising model and the realizing method are introduced. The GIS realized the map layers management for complex spatial data efficiently. Finally a real case is studied and it is demonstrated that the method is feasible and has good potential application.

Support Vector Machine Model of Coal Floor Water Inrush Forecast
This content is only available via PDF.
You do not currently have access to this chapter.
Close Modal

or Create an Account

Close Modal
Close Modal