STATISTICS - Temporal and spatial data: handling non-independence
| When |
Apr 19, 2023
from 01:15 PM to 04:45 PM |
|---|---|
| Where | room 200, Tennenbacherstraße 4 |
| Add event to calendar |
vCal
iCal
|
This workshop is part of a series (1-4) and addresses PhD students with an advanced level of statistics. Each session can be booked independently.
Question
Temporal and spatial data: handling non-independence
Returning, full circle, to the prediction-inference-dichotomy, we look at why non-independence is/not a problem for our analysis.And if we decide it is, what to do about it.
Central themes will be independence diagnostics, Generalised Least Squares and the customised variance-covariance matrix of the response, but we will also look at some approaches that scale to large data and machine learning approaches. And how to validate them.
Facilitator: Prof. Carsten Dormann (Chair of Biometry and Environmental System Analysis)
Number of participants: minimum 6 - maximum 10
Registration deadline: April 12