CodPy : A Python Library for Machine Learning, Mathematical Finance, and Statistics

This monograph presents the algorithms that are implemented in the Python library CodPy—an acronym that stands for the “Curse Of Dimensionality in PYthon”. This library provides the user with a support vector machine (SVM) and encompasses a broad set of applications. The proposed algorithms apply to partial differential equations arising, for instance, in mathematical finance and fluid dynamics, as well as to discrete models arising in machine learning and statistics. Our basic numerical strategy relies on a combination of ideas from the theory of reproducing kernel Hilbert spaces (RKHS) and the theory of optimal transport. The authors have developed this library over the past decade, originally for applications in mathematical finance, and it has now reach a stage where it can be applied to problems of interest to engineering and industry