BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.4.0.1//EN TZID:Europe/Paris X-WR-TIMEZONE:Europe/Paris BEGIN:VEVENT UID:394@isdm.umontpellier.fr DTSTART;TZID=Europe/Paris:20250203T140000 DTEND;TZID=Europe/Paris:20250203T140000 DTSTAMP:20260828T114650Z URL:https://isdm.umontpellier.fr/events/personalizing-treatment-with-causa l-inference-and-scalably-evaluating-llms-in-medicine-3/ SUMMARY:Personalizing Treatment with Causal Inference and Scalably Evaluati ng LLMs in Medicine DESCRIPTION:Online\n\nMachine Learning in Montpellier\, Theory &\; Pract ice\n\nThis talk examines two critical aspects of data-driven medicine: pe rsonalized treatment strategies using causal inference and the robust eval uation of large language models (LLMs) for clinical applications. In the f irst part\, we present a novel approach to personalized medicine\, applyin g causal statistical learning to observational data to develop individuali zed treatment rules. We focus on optimizing the timing of renal replacemen t therapy initiation in acute kidney injury\, demonstrating: (i) the estim ation and validation of an optimal dynamic strategy\, and (ii) a comprehen sive framework for evaluating individualized rules using observational dat a. In the second part\, we tackle the challenge of evaluating LLMs in medi cine\, focusing on the generation of hospital course summaries. Current ev aluation methods are often either unscalable (physician-led) or untrustwor thy for clinical settings (LLM-as-a-judge). We propose a rubric-based appr oach to LLM evaluation that combines the scalability of automated methods with the trustworthiness demanded by medical applications\, paving the way for responsible deployment of LLMs in healthcare.\n\nMachine Learning in Montpellier\, Theory &\; Practice ATTACH;FMTTYPE=image/jpeg:https://isdm.umontpellier.fr/wp-content/uploads/ 2026/06/ml-mtp-gC78d5.png CATEGORIES:ML MTP END:VEVENT BEGIN:VTIMEZONE TZID:Europe/Paris X-LIC-LOCATION:Europe/Paris BEGIN:STANDARD DTSTART:20241027T020000 TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET END:STANDARD END:VTIMEZONE END:VCALENDAR