Parallel machine learning

[curiosity]

Hi all and prof. Yaser,
Recently we started to see several parallel machine learning frameworks like graph lab, systemml, and graphci etc. I really wonder what you guys feel like whether there is really a need for a general framework because it seems like many companies develop their in house solutions for their machine learning tasks. Especially I really wonder what prof. Yaser thinks about these frameworks..

Thanks

[yaser]

Originally Posted by curiosity

Recently we started to see several parallel machine learning frameworks like graph lab, systemml, and graphci etc. I really wonder what you guys feel like whether there is really a need for a general framework because it seems like many companies develop their in house solutions for their machine learning tasks. Especially I really wonder what prof. Yaser thinks about these frameworks.

The honest answer is that it is too early to tell. Big Data is still a maturing field, and many of the techniques (from ML and from other disciplines) that are being tried are still in the exploratory phase. This makes it an exciting time from a research point of view.