Scala and Machine Learning

Martin Odersky didn't initially expect Scala to find a following in the field of machine learning because of machine learning's large appetite for memory and numeric computation. But the field is expanding in new ways, with interest in parallel and distributed computation, dynamically changing model structures, and the desire to put easy-to-use DSLs into the hands of non-experts. This talk will describe these trends and discuss several machine learning projects that use Scala, including FACTORIE, a 30k-line DSL for graphical models whose development is being sponsored by Google and the NSF.