BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Model Predictive Scoring shows specific BAO points (not Supernovae
 ) drive the dynamic dark energy tension
DTSTART;VALUE=DATE-TIME:20261028T175500Z
DTEND;VALUE=DATE-TIME:20261028T180500Z
DTSTAMP;VALUE=DATE-TIME:20261026T031956Z
UID:indico-contribution-3630@nyx.physics.mcgill.ca
DESCRIPTION:Speakers: Tanveer Karim (University of Toronto)\nThe recent an
 alyses of DESI DR2 BAO\, Planck CMB and various Supernovae datasets have s
 hown preferences for dynamic dark energy (DDE) over cosmological constant 
 at various levels of significance\, under frequentist model comparison tes
 ts. However\, the same analysis done with a pure Bayesian model comparison
  test can lead to the opposite conclusions. In contrast to these tradition
 al model comparison metrics\, we approach the problem along a third axis -
 - which model better predicts unseen data. We use the Expected Log Predict
 ive Density (ELPD) metric to score the predictiveness of cosmological cons
 tant and DDE models\, using the leave-one-redshift-block-out cross-validat
 ion estimator. Aggregated\, we find modest preferences for DDE\, primarily
  driven by the BAO. Decomposing the ΔELPD score by redshift we find that 
 a single BAO tracer\, LRG2 (z=0.706)\, supplies the entire DDE preference.
  Interestingly\, the biggest outlier point\, LRG1\, contributes little to 
 the model predictive scoring because both models fail to predict the obser
 vation by equal amount. On the other hand\, all SNIa datasets' ΔELPD scor
 es are consistent with 0\, indicating that at an out-of-distribution predi
 ctive level\, the DDE tension is entirely driven by one BAO point\, not SN
 Ia.\n\nhttps://nyx.physics.mcgill.ca/event/587/contributions/3630/
LOCATION:Online 
URL:https://nyx.physics.mcgill.ca/event/587/contributions/3630/
END:VEVENT
END:VCALENDAR