Online Workshop: Take your species distribution models to the next level with Bayesian nonparametric regression
Organizer: A. Márcia Barbosa
Date and time: 7-8 June 2021, 3-7 pm UTC (Note that many countries change to daylight saving time (UTC+1) just before the workshop dates)
Length of workshop: 8 hours, divided into two 4-hour blocks.
Maximum number of attendees: 10
Expected pre-experience: Basic experience with species distribution (or ecological niche) models and with running and modifying R scripts.
Summary: Bayesian Additive Regression Trees (BART) are a powerful machine learning technique with very promising applications in ecology and biogeography in general, and in species distribution modelling (SDM) in particular. BART can produce highly accurate predictions without overfitting to noise or to particular cases in the data. Notably, unlike most SDM methods, BART generally show a well-balanced performance regarding both main aspects of predictive accuracy, namely discrimination (i.e., distinguishing presence from absence localities) and calibration (i.e., having predicted probabilities reflect gradual occurrence frequencies). Moreover, the Bayesian framework inherently handles uncertainty and has a built-in complexity penalty with very sensible defaults. In this workshop we will see the advantages of this method and produce robust distribution models, identify relevant predictors, map prediction uncertainty, plot partial dependence curves with Bayesian credible intervals, and map relative presence probability regarding particular predictors. Reproducible R scripts and example data will be provided for all analyses. Participants may also use their own species presence-(pseudo)absence data and predictor variables, but keep in mind that computation time can be very large for large datasets.
Materials needed: A computer with recent versions of R and RStudio, plus R packages ’embarcadero’, ‘blockCV’ and ‘modEvA’ installed; A good enough internet connection for live video sessions, and preferably a webcam for enhanced interactivity.
Cost: 120 euros