Microsoft has purchased Revolution (an R company). Here are some reactions:
- http://www.r-bloggers.com/microsoft-acquires-revolution-analytics-news-roundup/
- http://www.forbes.com/sites/danwoods/2015/01/27/microsofts-revolution-analytics-acquisition-is-the-wrong-way-to-embrace-r/
- http://www.forbes.com/sites/danwoods/2015/01/28/how-microsoft-could-implement-a-hybrid-embrace-of-r/
- http://blogs.wsj.com/cio/2015/01/26/microsoft-courts-data-scientists-with-revolution-analytics-buy/
- http://www.infoworld.com/article/2876535/application-development/can-microsoft-make-r-easy.html
Some thoughts:
- MS has already R inside Azure-ML. https://studio.azureml.net/
- I always thought of ML (Machine Learning) as a somewhat smallish but highly fashionable subset of Statistics, but now it seems ML is hyped so much to completely obscure Statistics.
- Is ML really that big?
- ML has little bit more of a CS and IT background (traditional statistics more of a math background with perhaps most applications in the social sciences?). May be CS/IT people are much better in their public relations. Also many companies nowadays see IT as core business, so may be there is more funding available for doing significant ML applications.
- MS has a poor performance record wrt analytics: poor Solver Foundation was abandoned in a bad way leaving some clients hanging to dry.
- MS already has a popular statistics package (it is called Excel). (Joking!)
- MS is a little bit late: IBM purchased SPSS in 2009.
- R is very successfully riding this hype cycle. Not bad for a somewhat funky and older language (rooted in S).
- Do we get an R.NET?
- TIBCO offers a commercial re-implementation of R.
- ML vs Statistics: http://brenocon.com/blog/2008/12/statistics-vs-machine-learning-fight/