RVM-BASED Ore Grade Forecasting Model and Its Application

摘要：

In order to establish 3D solid model in geological fields, the key question is to obtain complete ore grade attribute data. Traditional forecasting methods such as neural networks, support vector machine (SVM) are adopted frequently. However, these methods are lack of necessary probability information and can not acquire the uncertainty of forecasts. In this paper, a new forecasting model is proposed based on relevance vector machine (RVM). Compared with other methods, the mistake rate and time complexity of RVM is lower and RVM has not any restriction on the selection of kernel function. Test results show that the proposed method has superior non-linear forecasting ability, higher precision of prediction and broad application prospects.

关键词： RVM information forecast 3D modeling ore grade

作者: WU HuiXin WANG Feng

作者单位: Dept.of Information Engineering North China University of Water Conservancy & Electric Power Zhengzhou, China

会议类型: 国际会议

会议名称: 2010 3rd IEEE International Conference on Computer Science and Information Technology(第三届IEEE计算机科学与信息技术国际会议 ICCSIT 2010)

会议地点: 成都

会议语种:英文

页码: 449-452

在线出版日期: 2010-07-07（万方平台首次上网日期，不代表论文的发表时间）