You are not currently logged in.
Access JSTOR through your library or other institution:
If You Use a Screen ReaderThis content is available through Read Online (Free) program, which relies on page scans. Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Predictive Mean Square Error and Stochastic Regressor Variables
Subhash C. Narula
Journal of the Royal Statistical Society. Series C (Applied Statistics)
Vol. 23, No. 1 (1974), pp. 11-17
Stable URL: http://www.jstor.org/stable/2347047
Page Count: 7
You can always find the topics here!Topics: Statism, Modeling, Gaussian distributions, Estimators, Least squares, Parametric models, Mathematical problems, Multiple regression, Regression coefficients, Statistical variance
Were these topics helpful?See something inaccurate? Let us know!
Select the topics that are inaccurate.
Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Preview not available
When prediction is the main objective, predictive mean square error (p.m.s.e.) seems to be a more reasonable criterion. Here we consider two approaches to improve the p.m.s.e. of the predicted response when predictor variables are stochastic and, in particular, follow a multivariate normal distribution. The first technique, the subset approach, uses only a subset of the available predictor variables to predict the response. A decision rule to select the subset is given. In the second method, the lambda approach, the regression coefficients are scaled down by a suitable constant. An estimator of the constant is suggested. Both techniques are illustrated by an example.
Journal of the Royal Statistical Society. Series C (Applied Statistics) © 1974 Royal Statistical Society