talk · community record
Guerrilla data-mining with Scala, Scoobi and R
The talk will cover my experience in using Scala to develop MapReduce programs on Hadoop at eBay: motivations, challenges and lesson learned. I will describe how we started using Java-based libraries on hadoop (basic MapReduce and then cascading) and finally moved to Scoobi recently. I will also discuss when to use Scala and when to use R for machine learning/data minining and why.