by Joseph Rickert
I always think of Strata Hadoop World and Predictive Analytics World as initiating the Spring conference season here in the San Francisco Bay Area. The rainy season is usually over by the end of March and it is a perfect time to visit. If you are traveling to either of these conferences from out of town and you are an R aficionado please arrange your plans to include one one or both of the Bay Area useR Group, BARUG, meetings that will be held in conjunction with these events.
For the past few years, the folks at O'Reilly and the organizers of PAW have very generously arranged for us to hold meetings on-site at the conference locations. This year, even though the two BARUG meeting will be held within a week of each other, we have manged to assemble a panel of speakers that I expect will rival the line up at the conferences themselves.
Our San Jose, Strata meeting will be held on March 29th. David Smith will begin the evening by describing how R is being integrated into the warp and woof of Microsoft's offerings. Hadley Wickham will present his ideas about managing many models with the help of the tidyr, purrr and broom R packages and H20's Erin LeDell will follow up on Hadley's theme and talk about model management with model stacking or “super learning” as it is sometimes called. This recent video of Hadley speaking at the WOMBAT conference in February covers some of the themes that he will be presenting.
We have assembled an even more diverse and eclectic line up of speakers for our April 4th Predictive Analytics World meeting in San Francisco. Keith Everett will present the GLMNet model behind his Oscar Predictions model. Then Megan Price, a long-time member and supporter of BARUG, will present the statistical techniques and R packages that the Human Rights Data Analysis Group, HRDAG, uses to compensate for the missing data problems that plague attempts to estimate conflict related casualties in Syria. This will be Megan's first BARUG talk since becoming Executive Director at HRDAG.
After Megan, Blair Hull, legendary trader, founder and managing partner of Ketchum Trading, will talk about market timing, big data and machine learning. Blair is all about algorithms and Ketchum Trading was a founding member of the R Consortium.
Max Kuhn, co-author of Applied Predictive Modeling and frequent BARUG speaker, will close out the evening with a talk on rule based regression models.
If you are planning on attending the meetings please register as soon as you can . Space will be limited.
For details and registration for the the San Jose, Strata meeting look here, and for the San Francisco PAW meeting go here. Also note that both Strata and PAW are offering conference discounts for BARUG members. The discount details are on the meeting sites linked to above.
Finally, if you haven seen the “Predict This” the dance video by PAW's Eric Siegel you can find it here.