Large sky surveys such as the ongoing Dark Energy Survey (DES) and the upcoming Large Synoptic Sky Survey (LSST) will generate enormous amounts of high quality imaging data of galaxies. These data, in many different filters, will provide invaluable information on galaxy evolution. The exploitation of these vast datasets will require advanced machine learning techniques. The student will build and test a convolutional neural networks based technique to classify the morphological types of galaxies in sky survey images automatically. The neural networks will be trained on images classified in the galaxy ZOO citizen science project and the technique will be tested on SDSS, Pan-Starrs, and/or DES images.