Chapter 10. Using Spark SQL in Deep Learning Applications
Deep learning has emerged as a superior solution to several difficult problems in machine learning over the past decade. We hear about deep learning being deployed across many different areas, including computer vision, speech recognition, natural language processing, audio recognition, social media applications, machine translation, and biology. Often, the results produced using deep learning approaches have been comparable to or better than those produced by human experts.
There have been several different types of deep learning models that have been applied to different problems. We will review the basic concepts of these models and present some code. This is an emerging area in Spark, so even though there are several different libraries available, many are in their early releases or evolving on a daily basis. We will provide a brief overview of some of these libraries, including some code examples using Spark 2.1.0, Scala, and...