Materials science, the discipline of finding materials well-suited for their purpose, is an important field whose results are used in vehicles, solar cells, and fusion reactors, among others. This field is traditionally based on experiments and computational methods, like Molecular Dynamics simulations. These methods, however, take a long time and limit the speed at which new materials are discovered. For this reason, Machine Learning based methods are on the rise, promising to give results at a fraction of the time, without sacrificing the quality of the results. These algorithms seek patterns in data gathered on different materials in the past to then make predictions about the properties of other, related compounds, or even directly generate new candidates. This thesis gives a short overview of how Machine Learning models work, and then shows how the technology is used for different material types to speed up the discovery of new materials.