Application of Hybrid Modeling to Air Quality Data by Combining CART Analysis with Regression Model

スポンサーリンク

概要

• 論文の詳細を見る

• Data mining is strategy and method for the exploration of the optimum model that exists in data set. The paper proposes a strategy methodology for data mining, called hybrid modeling, when exploring the optimal model to be applied to an air quality data. The modeling technique combines a non-parametric model, Classification And Regression Trees (CART), with a parametric model such as multiple regression analysis. The modeling can improve the accuracy of individual modeling by CART analysis or regression modeling. The modeling can handle any case replacing missing value for any predictor with surrogate CART variable. Autocorrelation of the residual time series and its residual mean square error after hybrid modeling can be improved by the application of an autoregressive model.

• 明治大学の論文

著者

• Ootaki Atsushi

  Department Of Mechanical System Science And Engineering School Of Science And Technology Meiji Unive

• KADOWAKI Tetsuo

  Department of Mechanical System Science and Engineering School of Science and Technology, Meiji University

関連論文

• Application of Hybrid Modeling to Air Quality Data by Combining CART Analysis with Regression Model

スポンサーリンク