Weighted Empirical Risk Minimization: Transfer Learning based on Importance Sampling

Robin Vogel, Mastane Achab, Charles Tillier, Stéphan Clémençon

Keywords: statistical learning theory, importance sampling, positive unlabeled (PU) learning, selection bias

TL;DR: When training and testing distributions are different, importance sampling works for many common practical cases.

Abstract: We consider statistical learning problems, when the distribution $P'$ of the training observations $Z'_1,\; \ldots,\; Z'_n$ differs from the distribution $P$ involved in the risk one seeks to minimize (referred to as the \textit{test distribution}) but is still defined on the same measurable space as $P$ and dominates it. In the unrealistic case where the likelihood ratio $\Phi(z)=dP/dP'(z)$ is known, one may straightforwardly extends the Empirical Risk Minimization (ERM) approach to this specific \textit{transfer learning} setup using the same idea as that behind Importance Sampling, by minimizing a weighted version of the empirical risk functional computed from the 'biased' training data $Z'_i$ with weights $\Phi(Z'_i)$. Although the \textit{importance function} $\Phi(z)$ is generally unknown in practice, we show that, in various situations frequently encountered in practice, it takes a simple form and can be directly estimated from the $Z'_i$'s and some auxiliary information on the statistical population $P$. By means of linearization techniques, we then prove that the generalization capacity of the approach aforementioned is preserved when plugging the resulting estimates of the $\Phi(Z'_i)$'s into the weighted empirical risk. Beyond these theoretical guarantees, numerical results provide strong empirical evidence of the relevance of the approach promoted in this article.

Code: https://drive.google.com/drive/folders/1-tWJ4n4WyXuTza8dLPngyHSVprKUZFVJ