This is our first Apache Spark Meetup in Munich. In this Meetup we will explain some Spark basics and show a small live demo in the first Talk by Danny. The second talk will be about implementing a ML based answer scoring with Spark and MLLib.
Between and after the talk we have time for drinks, food and conversations.
! The talks will be in German !
"Spark Basics - RDD, SQL, Mllib, GraphX" - Danny Linden
"With the rapid adoption of Apache Spark—one of the most active Apache projects today—and the need for programs to solve the world's greatest problems, distributed computing has resurfaced as a hot commodity that can take your career to the next level. More importantly, Spark opens the door to some really cool and impactful applications. Spark is a leap forward in distributed computing, allowing you to perform faster and more complex analyses on your Hadoop cluster and in the cloud. This presentation will give a short introduction to basic Spark concepts such as RDDs, transformations, actions, and executors. We will also cover recent developments in the Spark community with DataFrames, SQL on Spark, GraphX."
"Speed Up Your Spark Job" - Christian Dedié
gutefrage.net is using spark extensively for Maschine Learning, BI and realtime processing of user behavior. This talk is about our learnings and pitfalls when implementing a ML based answer scoring (ordering) for all questions on gutefrage.net. E.g. improve reliability and throughput of spark jobs with read/write access to relational datasources, or optimize HDFS based data structures for best performance. Starting with Dataframes and Spark SQL, we experienced some major improvements when implementing the same functionality based on RDDs.
About Christian Dedié:
Christian Dedié has 20 years of experience as a software engineer. He's a passionate Scala developer and Continuous Delivery advocate. In the last years he focused on big data projects using Polyglot Persistence and Maschine Learning. He is co-founder of the open source project "Flyway - Database Migrations Made Easy".