Strata NYC Reporting: Monday @ Big Data Camp, Tuesday @ Strata Retrospective
This is Russell Jurney, your Big Data reporter on the ground here at Strata NYC/Hadoop World at the New York Hilton. Monday night’s main event was Big Data Camp. As in any unconference, the best action was in the hallway, meeting people you only know by reputation or from twitter. Highlights were:
- Microsoft’s demonstration of Excel -> Power Pivot -> Hortonworks Data Platform
- Druid: Real-Time Analytics at a Billion Rows Per Second by Eric Tschetter, Co-founder of Metamarkets
- The Horrors Hidden in Your Models by Steven Hillion
In light of today’s announcement – the Hadoop market just got MUCH bigger 
In-RAM stores are an interesting new development as RAM becomes cheaper and cheaper, and can augment a Hadoop-centric workload.
This talk stressed the importance of unit testing your statistical models and finding areas where they fall-over, then working with customers to understand the problem. A humorous use-case involving a hoax ‘finger-in-chili’ incident was examined.
Tuesday’s tutorial sessions were great. My favorites were:
- A Hands-on Introduction to Cross-disciplinary Analytics With Python
- Hadoop Data Warehousing with Hive
This used iPython notebooks and scikit-learn to teach applied statistics.
This session featured live Hadoop clusters with Hive installed for users to bang on. Love the interactivity and forethought!
Check back tomorrow for coverage of Wednesday’s technical sessions!