H2O, an open source in-memory machine learning and predictive analytics company for big data, recently announced that MapR Technologies has certified H2O product for Apache Hadoop. The certification implies that MapR clients can run open source H2O advanced algorithms on data stored in Hadoop clusters without the need for data transfers.
H2O is extensible and users can build blocks using simple math legos in the core. H2O keeps familiar interfaces like R, Excel & JSON so that big data enthusiasts and experts can explore, munge, model and score data sets using a range of simple to advanced algorithms.
H2O recently announced that its flagship H2O product is available on the Intel (News - Alert)® Distribution for Apache Hadoop. Companies that use the Intel Distribution are now able to use open source H2O to run advanced algorithms on existing data stored in Hadoop clusters with no need for data transfers. By combining the power of H2O's highly predictive algorithms with the high performance Intel Distribution, organizations can discover valuable insights up to 100x faster than alternatives. The key to H2O's interactive performance is its fast in-memory parallel processing. Cache oblivious implementations of algorithms over columnar compressed data delivers distributed machine learning algorithms at blazing speeds.
“Together MapR and H2O are working to provide enterprises with easy access to the fastest predictive analytics on the sturdiest Hadoop distribution,” said Bill Bonin, vice president of business development of MapR Technologies. “We’re driving greater value to our customers through a variety of popular use cases from fraud detection and attribution analysis, to analyzing traffic patterns.”
“H2O brings near real-time modeling and low-latency scoring to MapR Enterprise Hadoop Distribution,” said SriSatish Ambati, co-founder and CEO of H2O. “Some of the first to realize increased ROI on their Hadoop investments have been Insurance and Financial Services companies through the use of pricing engines and fraud prediction.”