Prediction of Water Inrush from Coal Floor Based on Small Sample Data Mining Technology and Realization Using MATLAB
Prediction of Water Inrush from Coal Floor Based on Small Sample Data Mining Technology and Realization Using MATLAB
复制标题
DOI：
10.4304/jsw.7.10.2241-2246
复制
发表时间：
2012-01
期刊：
影响因子：
--
通讯作者：
Li Yang-;Chengcheng Liu;Jing Han;W. Sheng
中科院分区：
文献类型：
--
作者：
Li Yang-;Chengcheng Liu;Jing Han;W. Sheng
在煤层底板突水样本较少的情况下，如何在样本数据有限的情况下挖掘更广泛的信息，是提高煤层底板突水预测精度的关键。为此，建立了基于相关分析和支持向量机的突水预测模型。该模型不仅对影响煤层底板突水的指标进行了简化，构建了煤层底板突水的指标体系，而且挖掘了样本数据的值，解决了小样本、非线性预测的问题。通过实证分析，利用MatLab建立的模型能够准确预测煤层底板是否发生突水。
For there are few samples of water inrush from coal floor, how to dig wider information under the circumstance of limited sample data is to improve the prediction accuracy of water inrush form coal floor. Therefore, prediction model of water inrush is established based on correlation analysis and support vector machine (SVM). Not only does the model simplify the influenced indexes to construct the index system of water inrush from coal floor, but also dig the value of sample data to solve the problem of small sample and nonlinear prediction. Through the empirical analysis, the model using MATLAB can accurately predict whether water inrush from coal floor occurs.