文档介绍：------------------------------------------------------------------------------------------------ ——————————————————————————————————————机器学习工具 WEKA 的使用总结, 包括算法选择、属性选择、参数优化一、属性选择: 1 、理论知识: 见以下两篇文章: 数据挖掘中的特征选择算法综述及基于 WEKA 的性能比较_ 陈良龙数据挖掘中约简技术与属性选择的研究_ 刘辉 2、 weka 中的属性选择 评价策略( attribute evaluator ) 总的可分为 filter 和 wrapper 方法,前者注重对单个属性进行评价,后者侧重对特征子集进行评价。 Wrapper 方法有: CfsSubsetEval Filter 方法有: CorrelationAtt ributeEval Wrapper 方法: (1) CfsSubsetEval 根据属性子集中每一个特征的预测能力以及它们之间的关联性进行评估, 单个特征预测能力强且特征子集内的相关性低的子集表现好。 Evaluates the worth ofa subset of attributes by considering the individual predictive ability of each feature along with the degree of redundancy between of features that are highly correlated ------------------------------------------------------------------------------------------------ —————————————————————————————————————— with the class while having low intercorrelation are preferred. For more information see: M. A. Hall (1998). Correlation-based Feature Subset Selection for Machine Learning. Hamilton, New Zealand. (2) WrapperSubsetEval Wrapper 方法中,用后续的学习算法嵌入到特征选择过程中,通过测试特征子集在此算法上的预测性能来决定其优劣, 而极少关注特征子集中每个特征的预测性能。因此, 并不要求最优特征子集中的每个特征都是最优的。 Evaluates attribute sets by using a learning scheme. Cross validation is used to estimate the accuracy of the learning scheme for a set of attributes. For more information see: Ron Kohavi, e H. John (1997). Wrappers for feature subset Intelligence. 97(1-2):273-324. Filter 方法: 如果选用此评价策略,则搜索策略必须用 Ranker 。(1) CorrelationAttributeEval 根据单个属性和类别的相关性进行选择。 Evaluates the worth of an attribute by measuring the correlation (Pearson's) between it and the class. ------------------------------------------------------------------------------------------------ —————————————————————————————————————— Nominal attributes are considered ona value by value basis by treating each value as an indicator. An overall correlation for a nominal attribute is arrived at via a weighted average. (2) GainRatioA