Classification of spontaneous combustion tendency grade for sulfide ores based on Gaussian process

摘要：

Classifying the spontaneous combustion tendency grade of sulfide ores is a highly complicated nonlinear problem. With the Gaussian process theory, this paper selects four indexes including main mineral and content, constant mean of oxidation rate, selfheating point and spontaneous combustion point as the characteristic factors in Gaussian process classification (GPC), and establishes the GPC model for classifying spontaneous combustion tendency grade of sulfide ores by supervising and training the experimental data of thirteen typical ore samples. Under the transcendental hypothesis from Gaussian process and the target of maximizing the posteriori probability, the model educes predicted value and related probability of four new samples data. The conclusion demonstrates that the classifying capability of GPC model is high performance and easily realized, which indicates the method has a promising application prospect in classification of spontaneous combustion tendency grade of sulfide ores.

关键词： sulphide ores spontaneous combustion tendency grade classification Gaussian process machine learning

作者: Shao liangshanl Li jianying Qiu yunfei Bai yuan

作者单位: Institute of Systems Engineering. College of Science ofLiaoning Technical University, Huludao, China School of Software. College of Science ofLiaoning Technical University, Huludao, China

会议类型: 国际会议

会议名称: International Symposium on Emergency Management 2011(ISEM‘2011)(第六届国际应急管理论坛暨中国(双法)应急管理专业委员会第七届年会)

会议地点: 北京

会议语种:英文

页码: 598-601

在线出版日期: 2011-11-19（万方平台首次上网日期，不代表论文的发表时间）