It's great to see so many in attendance for the first-ever Predictive Analytics World. I've heard a lot of interesting talks already (too bad some are in parallel making choices hard). A recap and highlights of the talks I was able to hear and the conversations I had during breakouts follows.
Eric Siegel's opening talk was a good kick-off and could have been entitled "Predictive Analytics for Pun and Profit." John Elder presented several noteworthy case studies and generated a lot of interest in the forthcoming book, Statistical Analysis & Data Mining Applications. Usama Fayyad's keynote on "New Challenges in Predictive Analytics: Social Networking, Direct Response Marketing and Understanding Customer Behavior" made it clear there are *many* exciting problems yet to be addressed. The track on product recommendations and the Netflix Prize provided a nice overview of recommender systems as well as some creative visualizations.
I've participated in some interesting conversations about poor data quality, the continued over-reliance on spreadsheets for business-critical functions, the latency of decisions due to manual model deployment processes, and the interest in analytic centers of excellence. I've also heard some nice comments on the plans for SAS to interface with R.
Looking forward to more tomorrow.