Methodology: Remove one observation. Training the rest of data that are sampled without replacement and given this observation's input, predict the response back. Replicate this N times and for each response, take a sample from these replicates with replacement. Average each responses of sample and again replicate this step N time for each observation. Approximate these N new responses and generate another N responses y'. Training these y' and predict to have N responses of each testing observation. The average of N is the final prediction. Each observation will do the same.
| Version: | 0.1 |
| Depends: | R (≥ 3.2.5), rpart , parallel , epandist , triangle, caret |
| Suggests: | MASS |
| Published: | 2018-05-09 |
| Author: | Moshu Xie |
| Maintainer: | Moshu Xie <mxie622 at gmail.com> |
| License: | GPL-2 |
| NeedsCompilation: | no |
| CRAN checks: | SQB results |
| Reference manual: | SQB.pdf |
| Package source: | SQB_0.1.tar.gz |
| Windows binaries: | r-devel: SQB_0.1.zip, r-release: SQB_0.1.zip, r-oldrel: SQB_0.1.zip |
| OS X binaries: | r-release: SQB_0.1.tgz, r-oldrel: SQB_0.1.tgz |
Please use the canonical form https://CRAN.R-project.org/package=SQB to link to this page.