污水处理厂机器学习综合评价  

Machine Learning Comprehensive Assessment of Sewage Disposal Plants

作　　者：张维[1] 杨旭才[1] 陆晓春[1] 史道济[2] 

机构地区：[1]天津大学管理学院,天津300072 [2]天津大学理学院,天津300072

出　　处：《天津大学学报（社会科学版）》2008年第2期118-121,共4页Journal of Tianjin University：Social Sciences

基　　金：天津市科技发展计划基金资助项目(06YFGZGX05400)

摘　　要：对污水处理厂的运营情况进行综合评价,既可以找出现有污水处理厂存在的不足,指明改进方向和目标,又可为今后建立新厂提供参考和借鉴。通过指标体系和随机线性评价模型层次分析法(analytical hierarchy process,AHP)权重,得到机器学习评价样本;采用随机森林、随机梯度Boosting和支持向量等六种机器学习方法和六种评价结果的平均值,对天津市14家污水处理厂运营情况进行排名。It is beneficial to comprehensively assess operation level of sewage disposal plants. The existing deficiency can be exposed, and the ameliorative direction and objective can also be shown. Based on indexes system, machine learning samples were designed in linear evaluation model by choosing analytical hierarchy process （AHP） weights at random. Fourteen sewage disposal plants in Tianjin were assessed by six machine learning algorithms, such as random forest, stochastic gradient boosting, support vector machine, etc. The final order were calculated according to the on average value of the 6 algorithms.

关 键 词：污水处理 综合评价 随机森林 随机梯度Boosting 

分 类 号：F207[经济管理—国民经济]