One of the original members of this meetup, Harlan Harris, will present his work on spatial analysis using modern tools.
Meetup organizers have to face a problem that's common to many other organizations. Where should I put my event, or factory, or chain of convenience stores, or cell phone tower? You could pay consultants many thousands of dollars to give you recommendations, or you can figure it out over a weekend using R! In this presentation, I will replicate an analysis we performed for DC Meetups, but using NYC home and work ZIP codes that YOU will provide (this information will be kept private). I'll show how to solve two location analysis problems: What would be the ideal location for NYOSP Meetup events? And what would be the best 3 locations that are convenient for everyone at least sometimes? [Sticking with downtown. -Jared] Techniques that will be discussed include: working with latitude/longitude data, constructing geometric cost functions, mapping, continuous optimization, global optimization, and dynamic report generation. R packages used include: knitr, plyr, reshape2, ggplot2, ggmap, and DEoptim.
Harlan Harris has a PhD in Computer Science (Machine Learning) from the University of Illinois at Urbana Champaign, and postdoctoral research experience in experimental psychology and cognitive science at Columbia University, UConn, and NYU. He currently works as a data scientist and architect for Sentrana, Inc., in Washington, DC. Harlan is the co-founder and co-organizer of the Data Science DC Meetup, is the cofounder and President of Data Community DC, and recently wrote a short e-book about data scientists called "Analyzing the Analyzers". Follow him on Twitter at @HarlanH.
Pizza starts at 6:30, Harlan begins at 7 and after we will head to a local bar.