基于信息增益优化支持向量机模型的煤矿瓦斯爆炸风险预测

Risk Prediction of Coal Mine Gas Explosion Based on Information Gain-Support Vector Machine(IG-SVM)Model

为了探索基于样本数据的煤矿瓦斯爆炸风险预测,依据本质安全理念构建了预测瓦斯爆炸风险的指标集,结合机器学习与特征优化算法提出了信息增益(information gain, IG)与支持向量机(support vector machine, SVM)的组合模型,通过对优化后的14种特征信息的分类学习,完成对风险未知样本的预测任务。以全国100家煤矿企业为研究对象,使用不同模型分别预测瓦斯爆炸风险并全面分析和比较,实验结果表明,经过IG优化后的SVM模型预测正确率达到了95.45%,相对于单一SVM模型提高了9.09%,同时高于其他预测模型,证明了该组合模型在瓦斯爆炸风险预测领域的优越性。

In order to explore the risk prediction of coal mine gas explosion based on the sample data,an index set for predicting gas explosion risk based on intrinsic safety concept was established.A combined model of information gain(IG)support vector machine(SVM)combined with machine learning and feature optimization algorithm was proposed,the prediction task of unknown risk samples was completed through the classification learning of 14 optimized feature information.100 coal mining companies across the country were selected as the research objects.Different models were used to predict the risk of gas explosion,and then comprehensively analyzed and compared.The experimental results show that the accuracy of SVM model after IG optimization reaches 95.45%.Compared with the single SVM model,it increases by 9.09%,higher than other prediction models.It proves the superiority of the combined model in the field of gas explosion risk prediction.