Dimension Reduction for Censored Regression Data via Sliced Average Variance Estimation

• Main Authors: Shih-Chia Liu, 劉士嘉
• Other Authors: Heng-Hui Lue
• Format: Others
• Language: en_US
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/86607285054930966633

碩士 === 東海大學 === 統計學系 === 94 === With censoring, high-dimensional regression becomes much more complicated. Since censoring can cause severe bias in estimation, modification to adjust such bias is needed to be made. Under the conditionally independent censoring condition, we propose the modification of sliced average variance estimation (SAVE) for estimating the joint effective dimension reduction (edr) space of lifetime and censoring time. Several simulation examples are reported and comparisons are made with the sliced inverse regression method of Li, Wang and Chen (1999).