SVM与BP神经网络在石煤提钒行业清洁生产评价中的对比研究
SVM and BP ANN applied to evaluation of cleaner productionin vanadium extraction from stone coal:a comparative study
投稿时间:2017-11-24  修订日期:2018-01-12
DOI:
中文关键词: 清洁生产  石煤提钒  支持向量机  BP神经网络  评价方法
英文关键词: extracting vanadium from stone coal  cleanerproduction  Support Vector Machine (SVM)  Genetic Algorithm (GA)  assessment methods
基金项目:湖北省自然科学基金(BZY16024)
作者单位E-mail
李佳 中南民族大学资源与环境工程学院 jiajiali1982@aliyun.com 
刘振宇 中南民族大学资源与环境工程学院  
摘要点击次数: 210
全文下载次数: 
中文摘要:
      为了比较BP神经网络和支持向量机2种机器学习算法在清洁生产评价中的应用能力,本研究首先从基础理论和应用原理入手,对比分析了2种机器学习算法的理论差异;其次,以石煤提钒生产工艺中水浸工艺为对象,完善了水浸工艺清洁生产评价指标体系,并以此为基础,对BP神经网络和支持向量机在清洁生产水平评价上做对比研究。结果表明,支持向量机方法分类精度为100%,BP神经网络分类精度为90%,且易陷入局部最优,因此支持向量机方法在解决小样本评价问题时具有较高的实用价值。
英文摘要:
      We evaluated the performance of two machine learning methods, BP artificial neural net (ANN) and support vector machine (SVM), for estimation of the application in cleaner production assessment. The theoretical analysis has been done from basic theory and application principle of these two methods first to compare the theoretical difference between them. Then, according to the water leaching process in vanadium extraction from stone coal, cleaner production assessment index framework in water leaching process has been developed. The performance of BP ANN and SVM, is comparatively analyzed in terms of cleaner production assessment. It demonstrated that theSclassificationSaccuracy of SVM reached 100%, theSclassificationSaccuracy of BP ANN was 90% and easy toSfall intoSSlocalS optimum. So, SVM method is more practical in small samples assessment.
View Fulltext   查看/发表评论  下载PDF阅读器
关闭