Maya Bercovitch - Apache Spark: Crunching Big Data Using Scala
15:30 - 16:20 Mild
Note: This talk will be given in Hebrew.
Big data ecosystems are everywhere nowadays, and it seems like everybody is trying to analyze data, develop algorithms and build models. In order to do that with big data, distributed processing becomes a must.
In this lecture I will introduce Apache Spark, a fast and general open source engine for large-scale data processing. With Spark, complex data processing tasks can be accomplished effectively using only a few lines in Scala. We will understand how Spark works and go over some great libraries that make data scientists smile.
Maya Bercovitch is a data science team leader at Supersonic - the promising Internet and mobile advertising startup company.
Maya's team is in charge of developing the algorithms responsible for maximizing performance and profits by predicting user behavior and personalizing ads. The team has recently conducted research for choosing a data processing engine optimized for achieving these purposes and has chosen to use Spark. The company's next generation machine-learning algorithms (as well as the entire back-end software) will be based on pure Scala.
Maya has been leading the development of big data prediction and classification algorithms for several years. Her M.Sc is in information system engineering, in which her research focused on machine learning and fraud detection.