頁籤選單縮合
題 名 | Regression Trees with Unbiased Variable Selection and Interaction Detection |
---|---|
作 者 | Loh,Wei-yin; | 書刊名 | Statistica Sinica |
卷 期 | 12:2 2002.04[民91.04] |
頁 次 | 頁361-386 |
分類號 | 319.51 |
關鍵詞 | Bagging; Bias correction; Bootstrap; Interaction detection; Piecewise linear; |
語 文 | 英文(English) |
英文摘要 | We propose an algorithm for regression tree construction called GUIDE. It is specifically designed to eliminate variable selection bias, a problem that can undermine the reliability of inferences from a tree structure. GUIDE controls bias by employing chi-secure analysis of residuals and bootstrap calibration of significance probabilities. This approach allows fast computation speed, natural extension to data sets with categorical variables, and direct detection of local two variable interactions. Previous algorithms are not unbiased and are insensitive to local interactions during split selection. The speed of GUIDE enables tow further enhancements-complex modeling at the terminal nodes, such as polynomial or best simple linear models, and bagging. In an experiment with real data sets, the prediction mean square error of the piecewise constant GUIDE model is within±20% of that of CART®. PIECWISE LINEAR guide models are more accurate; with bagging they can outperform the spline-based MARS® method. |
本系統中英文摘要資訊取自各篇刊載內容。