Deep learning topological phases of matter

This thesis is aimed at showing how to set up a typical problem of Condensed Matter physics in a Deep Learning framework. In order to do this we will introduce the Kitaev model (a superconducting quantum wire with topological properties) with nearest neighbor coupling, next to nearest neighbor coupling and an interacting term. Then we will present the Machine Learning techniques we are going to use. Finally we will apply them to train a Neural Network and a Convolutional Neural Network on recognizing the topological phases of matter of the non-interacting model to test it on the classification of interacting data.