Statistical Numerical Approximation

Abstract

Although numerical approximation and statistical inference are traditionally seen as entirely separate subjects, they are intimately connected through the common purpose of making estimations with partial information. This shared purpose is currently stimulating a growing interest in statistical inference/machine-learning approaches to solving PDEs While this interest might be perceived as a recent phenomenon, interplays between numerical approximation and statistical inference are not new. Indeed, they can be traced back to Poincar's course in probability theory (