期刊
ANNALS OF APPLIED STATISTICS
卷 2, 期 3, 页码 916-954出版社
INST MATHEMATICAL STATISTICS-IMS
DOI: 10.1214/07-AOAS148
关键词
Regression; classification; learning ensembles; rules; interaction effects; variable importance; machine learning; data mining
资金
- Department of Energy [DE-AC02-76SF00515]
- NSF [DMS-97-64431]
General regression and classification models are constructed as linear combinations of simple rules derived from the data. Each rule consists of a conjunction of a small number of simple statements concerning the values of individual input variables. These rule ensembles are shown to produce predictive accuracy comparable to the best methods. However, their principle advantage lies in interpretation. Because of its simple form, each rule is easy to understand, as it its influence on individual predictions, selected subsets of predictions, or globally over the entire space of joint input variable values. Similarly, the degree of relevance of the respective input variables can be assessed globally, locally in different regions of the input space, or at individual prediction points. Techniques are presented for automatically identifying those variables that are involved in interactions with other variables, the strength and degree of those interactions, as well as the indentities of the other variables with which they interact. Graphical representations are used to visualize both main and interaction effects.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据