Big Data Processing on Embedded Systems Date: This event took place live on August 25 2015 Presented by: Stefan Zapf Duration: Approximately 60 minutes. Cost: Free Questions? Please send email to Description:The webcast is a hands-on demonstration of the theory and practice of the popular HyperLogLog algorithm. Google created a version of HyperLogLog as part of their Powerdrill database. We will use it to build an embedded system that can count the distinct number of millions of data points with just a few kilobytes of memory. With the advent of the Internet of Things, data processing on embedded systems becomes increasingly important and sublinear algorithms such as HyperLogLog can be one building block in this new world. Based on the use case of counting distinct passengers at London subway station, you will learn:
About Stefan ZapfStefan Zapf has over twelve years of industry experience, and a specialization in algorithms and software architecture, Stefan has worked on projects ranging from game architecture, over developing sound localization algorithms for Rainforest Connection to consultancy for Redheads Ltd. In addition to his first loves — programming and computer science — Stefan has also served as a teacher at the Community College Bamberg, and has taken classes in law and social sciences at the University of Cambridge. Twitter: @daebwae You might also be interested in |
|
|