I believe that when people look back in 50 years and assess the first two decades of the century, deep learning will be at the top of the list of technical innovations. The theoretical foundations of deep learning were established in the 1950s, but it wasn’t until 2012 that the potential of deep learning became evident to nonspecialists. Now, almost a decade later, deep learning pervades our lives, from smart speakers that are able to seamlessly convert our speech into text to systems that can beat any human in an ever-expanding range of games. This book examines an overlooked corner of the deep learning world: applying deep learning to structured, tabular data (that is, data organized in rows and columns).