In this lab, you will learn how to train deep learning models for with supervised (image classification) and unsupervised (retail similarity) learning using H2O DeepWater platform on a GPUs. This hands-on lab will be done primarily in Python and H2O's native interface Flow and will use TensorFlow, MXNet, and Caffe. You will use pre-defined networks such as ResNet V2 along with custom networks. You will also compare and ensemble deep learning with other machine learning algorithms such as GLM and GBM and deploy models for inference. Prerequisites: Attendees should be familiar with machine learning and neural network basics. This lab utilizes GPU resources in the cloud, you are required to bring your own laptop.