The Risk Inflation Criterion for Multiple Regression
Foster, Dean P. ; George, Edward I.
Ann. Statist., Tome 22 (1994) no. 1, p. 1947-1975 / Harvested from Project Euclid
A new criterion is proposed for the evaluation of variable selection procedures in multiple regression. This criterion, which we call the risk inflation, is based on an adjustment to the risk. Essentially, the risk inflation is the maximum increase in risk due to selecting rather than knowing the "correct" predictors. A new variable selection procedure is obtained which, in the case of orthogonal predictors, substantially improves on AIC, $C_p$ and BIC and is close to optimal. In contrast to AIC, $C_p$ and BIC which use dimensionality penalties of 2, 2 and $\log n$, respectively, this new procedure uses a penalty $2 \log p$, where $p$ is the number of available predictors. For the case of nonorthogonal predictors, bounds for the optimal penalty are obtained.
Publié le : 1994-12-14
Classification:  Decision theory,  minimax,  model selection,  multiple regression,  risk,  variable selection,  62C99,  62J05,  62C20
@article{1176325766,
     author = {Foster, Dean P. and George, Edward I.},
     title = {The Risk Inflation Criterion for Multiple Regression},
     journal = {Ann. Statist.},
     volume = {22},
     number = {1},
     year = {1994},
     pages = { 1947-1975},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1176325766}
}
Foster, Dean P.; George, Edward I. The Risk Inflation Criterion for Multiple Regression. Ann. Statist., Tome 22 (1994) no. 1, pp.  1947-1975. http://gdmltest.u-ga.fr/item/1176325766/