This PhD Thesis concerns adaptative job scheduling and resource allocation. We target -among others- High Performance Computing machines. Specifically, we aim to develop machine learning techniques for these problems, in order to leverage the vast quantities of (meta-)data available today in these platforms.