Model free variable selection via sufficient dimension reduction
Duration: 20 mins 10 secs
Share this media item:
Embed this media item:
Embed this media item:
About this item
Description: |
Li, L (North Carolina)
Friday 27 June 2008, 09:40-10:00 Future Directions in High-Dimensional Data Analysis |
---|
Created: | 2008-07-14 14:40 | ||
---|---|---|---|
Collection: | Statistical Theory and Methods for Complex, High-Dimensional Data | ||
Publisher: | Isaac Newton Institute | ||
Copyright: | Li, L | ||
Language: | eng (English) | ||
Credits: |
|
Abstract: | Sufficient dimension reduction (SDR) has proven effective to transform high dimensional problems to low dimensional projections, while losing no regression information and pre-specifying no parametric model during the phase of dimension reduction. However, existing SDR methods suffer from the fact that each dimension reduction component is a linear combination of all the original predictors, and thus can not perform variable selection. In this talk, we propose a regularized SDR estimation strategy, which is capable of simultaneous dimension reduction and variable selection. We demonstrate that the new estimator achieves consistency in variable selection without requiring any traditional model, meanwhile retaining root-n estimation consistency of the dimension reduction basis. Both simulation studies and real data analyses are reported. |
---|
Available Formats
Format | Quality | Bitrate | Size | |||
---|---|---|---|---|---|---|
MPEG-4 Video | 480x360 | 1.84 Mbits/sec | 278.79 MB | View | Download | |
WebM | 480x360 | 407.53 kbits/sec | 60.24 MB | View | Download | |
Flash Video | 480x360 | 803.74 kbits/sec | 119.31 MB | View | Download | |
iPod Video | 480x360 | 504.98 kbits/sec | 74.96 MB | View | Download | |
QuickTime | 384x288 | 846.53 kbits/sec | 125.66 MB | View | Download | |
MP3 | 44100 Hz | 125.01 kbits/sec | 18.34 MB | Listen | Download | |
Auto * | (Allows browser to choose a format it supports) |