14
2
v1v2 (latest)

Optimal selection of sample-size dependent common subsets of covariates for multi-task regression prediction

Abstract

An analyst is given a training set consisting of regression datasets DjD_j of different sizes, which are distributed according to some GjG_j, j=1,,Jj=1,\ldots,\cal J, where the distributions GjG_j are assumed to form a random sample generated by some common source. In particular, the DjD_j's have a common set of covariates and they are all labeled. The training set is used by the analyst for selection of subsets of covariates denoted by P(n){P}^*(n), whose role is described next. The multi-task problem we consider is as follows: given a number of random labeled datasets (which may be in the training set or not) DJkD_{J_k} of size nkn_k, k=1,,Kk=1,\ldots,K, estimate separately for each dataset the regression coefficients on the subset of covariates P(nk){P}^*(n_k) and then predict future dependent variables given their covariates. Naturally, a large sample size nkn_k of DJkD_{J_k} allows a larger subset of covariates, and the dependence of the size of the selected covariate subsets on nkn_k is needed in order to achieve good prediction and avoid overfitting. Subset selection is notoriously difficult and computationally demanding, and requires large samples; using all the regression datasets in the training set together amounts to borrowing strength toward better selection under suitable assumptions. Furthermore, using common subsets for all regressions having a given sample size standardizes and simplifies the data collection and avoids having to select and use a different subset for each prediction task. Our approach is efficient when the relevant covariates for prediction are common to the different regressions, while the models' coefficients may vary between different regressions.

View on arXiv
Comments on this paper