I am new to Big-data and looking for a good platform to perform recommendations,clustering and classification. I understand Mahout has many algorithms to do this. Also R itself being a very good analytical tool is more than helpful for achieving this. Since I'm looking into bigdata I'll be using R+hadoop (via streaming/RHIPE/RHadoop etc.) if I decide to take this path. I'm seeking a comparison with respect to performance, scalability, ease of use, maturity etc considering that the difference in complexity of the language used (Java in mahout and R ) is not a factor to me at this point.
Asked
Active
Viewed 1,191 times
-1
-
If you want to use Hadoop, what is going to stop you from using Mahout as it runs on Hadoop? did you find any drawbacks? – Swamy Jun 18 '13 at 11:08
1 Answers
2
Mahout integrates many algorithms, and it deals with three questions (1) Recommendation (Collaborative filtering algorithm) (2) Clustering (3) Classification
R is just for statistic, so if you want to do these three tasks above, Mahout is a great tool.
BTW, Mahout is a Maven project, so you need to install Maven.
Freya Ren
- 2,086
- 6
- 29
- 39