779 Views
Track: Short Course
Drew Peltier
Northern Arizona University
Flagstaff, AZ, USA
Short courses are available to attendees who signed up in advance and paid an additional a short course fee. If you signed up for this short course, you should have received a direct link to it by email on Thursday.
This 6 hour short course includes a 30 minute break in the middle.
Ecologists increasingly rely spatially, temporally, or hierarchically variable data. Contemporary ecological problems require synthesis of multiple, often incomplete data sources, arising from mixtures of nonlinear and/or non-Gaussian processes. Hierarchical Bayesian statistical methods are powerful tools for analyzing disparate, large, and/or complex data sets.
While “canned” R packages are becoming available for performing traditional analyses within a Bayesian framework, a key advantage of Bayesian methods is custom model building. User-designed models enable flexible incorporation of experimental design, theory, and/or prior system knowledge. Creative, iterative model building is a key way to learn about the structure and functioning of your ecological system.
This short course covers introductory level Bayesian modeling. We will demonstrate the use of JAGS (freely available Bayesian software package), and discuss alternative model fitting software (e.g. Stan). Thus, participants should have some familiarity with R (e.g., for loops, list structuring, indexing).
Participants will develop and implement a Bayesian model based on a selection of data-motivated example problems. These examples will familiarize participants with different data structures or analysis techniques relevant for addressing ecological problems.
By the end of the short course, participants will understand the fundamentals of Bayesian modeling and implement basic hierarchical models. We will provide reference materials so participants can explore these topics in greater depth on their own. These materials should serve as a starting point for those interested in employing or further developing their skill in Bayesian methods.
Participants can access materials for this short course on GitHub.
The $35 fee for this short course includes access to the live session and any uploaded supporting documents during and after the conference.