3003 — Implementing High Value Spatial Analytics in the Context of VA Data.
Lead/Presenter: Evan Carey, COIN - Seattle/Denver
All Authors: Carey EP (Seattle/Denver COIN)
Glorioso TJ (Seattle/Denver COIN)
Understand how to leverage VA and non-VA spatial data to answer high-value questions. -Understand the sources of VA spatial data -Identify external spatial data sources that can be linked to VA data -Understand the general types of spatial data (point pattern versus areal) -Review computing technology to implement spatial analysis approved and available in the VA (hands on focus is on R) -Introduction to potential statistical methods to make spatial inference including basic Kriging, spatial GAM models, and hierarchical Bayesian models using R-INLA
The workshop will leverage simulated data (no PHI) that looks very similar to VA spatial data. We will begin with didactic overview of types and sources of spatial data in the VA. The majority of the workshop will involve reviewing types of spatial analytic approaches using the fake VA data. All participants will be provided with software installation instructions, along with the data and code files required to generate all of the course content. Attendees will be able to install all software ahead of time (free, open source) and then run and modify code during the workshop according to their comfort level. In addition to reviewing the fundamentals of spatial data and analytic approaches, we will facilitate a group discussion of what makes a spatially oriented analysis valuable as opposed to just interesting. The presenters of this workshop have extensive experience implementing spatial analytics in the VA; they will close with workshop with a discussion of their experience implementing spatial analytics in the VA to date.
Attendees to this workshop will be interested in developing a deeper understanding of what spatial analytics are possible in the VA environment. This workshop will be appropriate for both researchers interested in proposing and overseeing spatial studies, as well as analysts looking to improve their technical skills in implementing spatial analytics.
Assumed Audience Familiarity with Topic:
Attendees should be familiar with VA data sources in general. Technical experience writing statistical code is not required.