By Mike.Hallett-Oracle on Oct 28, 2015
Partners can differentiate their services by adding extra smart predictive and advanced analytics to Oracle BI deployment projects. Indeed OBI 12c now includes some capabilities integrated into OBI: but there are many additional techniques and tools to provide “data science” integrating Oracle Database, OBI and Hadoop with Advanced Analytics.
The core database Oracle Advanced Analytics product consists of two components, and you can learn more of these from a series of on-demand free of charge tutorials:
- ODM - Oracle Data Mining 4.1 OBE Series
- R – or more specifically ORE (open source R as deployed inside an Oracle database server) - Oracle R Enterprise v 1.4 - Tutorial Series
And there is a more general introduction @ Video Lecture Series Data Science Boot-Camp Tutorials. To get some practice on getting this all to work together, you can download pre-packaged VMs from OTN all set up ready to work:
- I recommend you Download the Demonstration VM for OBI 11g SampleApp v506 with Big Data … this has Advanced Analytics, DB12c, OBI and Cloudera Hadoop combined.
- If you want to also try our new “Big Data Discovery” then use this VM ~ Download VM Oracle Big Data Lite 4.2.1 Including Big Data Discovery
Oracle University also offers Data Mining Techniques, available as Training on Demand, which includes access to labs to do the practical - you can see a demo on this course here. There is also the Oracle R Enterprise Essentials course available as a scheduled Live Virtual Class.
There are many more aspects to “Advanced Analytics” with Oracle, for example you can follow some articles @ advanced and @ https://blogs.oracle.com/R/ and https://blogs.oracle.com/datamining/. Among the other components critical to specific types of analytical problems are the Spatial and the RDF-Graph options which run in the Database and we now have Big Data Spatial and Graph Analytics for Hadoop.
Oracle can even provide Real-Time Decisions using machine learning to make optimal decisions in the circa 100ms timeframe; reacting to events such as fraud detection, machine failure or to increase revenue by improving the probability of click-to-buy on a web-commerce shopping cart.