The images used to train this CNN are property of IUPUI so I can't upload them.
This uses an autoencoder to extract features from an image, feed them through a decision tree and output a crime statistic. The images used during the assignment were Google street view pictures of bars and liquor stores.