Hi Sameer, MLbase started out as a set of three ML components on top of Spark. The lowest level, MLlib, is now a rapidly growing component within Spark and is maintained by the Spark community. The two higher-level components (MLI and MLOpt) are experimental components that serve as testbeds for MLlib.
A prototype of MLI was released last fall, and since then many ideas from this prototype and the associated paper <http://www.cs.berkeley.edu/~ameet/mlinterface_icdm.pdf> have been (or are in the process of being) ported to MLlib. MLOpt, which ultimately aims to automate the task of ML pipeline construction, is an active area of research in the AMPLab. We are still exploring various research ideas, and given the nature of research, it's not clear when we will be ready to release any code. For more details check out mlbase.org, my talk <https://databricks-training.s3.amazonaws.com/slides/Spark_Summit_MLlib_070214_v2.pdf> at the Spark Summit about MLlib (slides 5, 6, 33, 34 describe MLbase), and Evan's slides <http://spark-summit.org/wp-content/uploads/2014/07/Ghostface-Towards-an-Optimizer-for-MLBase-Sparks-Talkwalkar-Franklin-Jordan-Kraska1.pdf> on the current status of MLOpt. -Ameet On Wed, Aug 27, 2014 at 12:52 PM, Sameer Tilak <ssti...@live.com> wrote: > Hi All, > I was wondering can someone please tell me the status of MLbase and its > roadmap in terms of software release. We are very interested in exploring > it for our applications. >
