BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The Spatial Confounding Environment - Dr Mauricio Tec\, Dept of Bi
 ostatistics\, Harvard University
DTSTART:20240305T160000Z
DTEND:20240305T170000Z
UID:TALK209713@talks.cam.ac.uk
CONTACT:Annabelle Scott
DESCRIPTION:Spatial confounding poses a significant challenge in scientifi
 c studies involving spatial data\, where unobserved spatial variables can 
 influence both treatment and outcome\, possibly leading to spurious associ
 ations. To address this problem\, we introduce SpaCE: The Spatial Confound
 ing Environment\, the first toolkit to provide realistic benchmark dataset
 s and tools for systematically evaluating causal inference methods designe
 d to alleviate spatial confounding. Each dataset includes training data\, 
 true counterfactuals\, a spatial graph with coordinates\, and smoothness a
 nd confounding scores characterizing the effect of a missing spatial confo
 under. It also includes realistic semi-synthetic outcomes and counterfactu
 als\, generated using state-of-the-art machine learning ensembles\, follow
 ing best practices for causal inference benchmarks. The datasets cover rea
 l treatment and covariates from diverse domains\, including climate\, heal
 th and social sciences. SpaCE facilitates an automated end-to-end pipeline
 \, simplifying data loading\, experimental setup\, and evaluating machine 
 learning and causal inference models. The SpaCE project provides several d
 ozens of datasets of diverse sizes and spatial complexity. It is publicly 
 available as a Python package\, encouraging community feedback and contrib
 utions.\n\nPreprint: https://arxiv.org/pdf/2312.00710.pdf.
LOCATION:Drum Building\, Madingley Rise Site\, West Cambridge and on zoom:
   https://zoom.us/j/6708259482?pwd=Qk03U3hxZWNJZUZpT2pVZnFtU2RRUT09
END:VEVENT
END:VCALENDAR