With Machine Learning now placed at the heart of many of applications, from recommender engines to trading bots and the growing popularity of Reactive Programming, we would like to present how to effectively combine the two. The focus of the presentation will be on online / streaming ML algorithms, which by working with live data streams are perfect for such domains as online trading or ad-hoc decision making.
Many such ML algorithms have already been implemented on top of Spark, but for truly reactive systems going to extremes (think IoT or HFT), deploying the whole related machinery (HDFS, Zookeeper, etc.) might be too heavy. Come and see how Akka Streams simplifies their implementation while also supporting them with its top tier performance. We will present real-life case studies for statistical and adversarial classes of algorithms in ad-hoc decision making, healthcare and finance.