Atomistic or ab-initio molecular dynamics simulations are widely used to predict thermodynamics and kinetics and relate them to molecular structure. A common approach to go beyond the time- and lengthscales accessible with such computationally expensive simulations is the definition of coarse-grained molecular models. Existing coarse-graining approaches define an effective interaction potential to match defined properties of high-resolution models or experimental data. In this paper we reformulate coarse-graining as a supervised machine learning problem. We use statistical learning theory to decompose the coarse-graining error and cross-validation to select to compare the performance of different models. We introduce CGnets, a deep learning approach, that learn coarse-grained free energy functions and can be trained by the force matching scheme. CGnets maintain all physically relevant invariances and allow to incorporate prior physics knowledge to avoid sampling of unphysical structures. We demonstrate that CGnets outperform the results of classical coarse-graining methods, as they are able to capture the multi-body terms that emerge from the dimensionality reduction.