Application of Tree-Based Methods on a Breast Cancer Data
碩士 === 靜宜大學 === 應用數學研究所 === 94 === When we want to know the characteristics of data, we analyze them first. After that, we can make a conclusion. In this study, we discuss a very popular method of statistics---Classification Tree. Besides the general greedy tree, Chipman、George and McCulloch in 1998...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2006
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/c9v4dx |
| Summary: | 碩士 === 靜宜大學 === 應用數學研究所 === 94 === When we want to know the characteristics of data, we analyze them first. After that, we can make a conclusion. In this study, we discuss a very popular method of statistics---Classification Tree. Besides the general greedy tree, Chipman、George and McCulloch in 1998 suggested that the method of Bayesian CART (Bayesian classification and regression tree) could apply into a breast cancer data and got the tree models. Now we use a same data to obtain new variables first from principal component analysis. Then utilizing the classification tree and cross-validation methods to analyze and discuss the conclusions and differences.
|
|---|