BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.4.2//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:0-334@lisn.upsaclay.fr
DTSTART;TZID=Europe/Paris:20251211T093000
DTEND;TZID=Europe/Paris:20251211T093000
DTSTAMP:20251202T111618Z
URL:https://www.lisn.upsaclay.fr/evenements/conformal-predictions-and-risk
 -control-in-machine-learning-models-to-improve-performance-and-human-decis
 ion-making/
SUMMARY:Conformal Predictions and Risk Control in Machine Learning Models t
 o Improve Performance and Human Decision-Making
DESCRIPTION:Keywords\nComputer vision\, uncertainty quantification\, confor
 mal prediction\, risk control\, deep learning\nAbstract\nMachine Learning 
 (ML) and Deep Learning (DL) models are increasingly deployed in critical a
 pplications where errors may lead to severe consequences. Despite their st
 rong predictive performance\, their inability to reliably express uncertai
 nty limits trust\, safety\, and compliance. Uncertainty Quantification (UQ
 ) addresses this challenge by estimating confidence levels in predictions\
 , enabling safer decision-making and improved performance.In this thesis\,
  I introduce post-hoc\, model-agnostic\, and black-box UQ methods applicab
 le to both classical tasks (regression\, classification) and structured ou
 tputs such as segmentation and object detection. The contributions include
 : (1) new conformal prediction techniques integrated into the MAPIE librar
 y\, including extensions to Gaussian processes\, and (2) risk-control appr
 oaches based on accuracy for biomedical detection\, as well as adaptive ri
 sk control for segmentation. Experiments on various real-world datasets de
 monstrate that these methods enhance the reliability and effectiveness of 
 machine learning systems.\nJury Composition\nMr. Mathieu SERRURIER\, Unive
 rsité Toulouse II Jean Jaurès – Reviewer\nMr. Lionel BOMBRUN\, Laborat
 ory for Integration from Material to System\, Université de Bordeaux – 
 Reviewer\nMs. Soundouss MESSOUDI\, Université de Technologie de Compiègn
 e – Examiner\nMr. Sébastien GERCHINOVITZ\, IRT Saint Exupéry – Exami
 ner\nMr. Antoine MANZANERA\, ENSTA-Paris\, IP Paris – Examiner\nMr. Gill
 es BLANCHARD\, Université Paris-Saclay – Examiner\nPublications\nAccess
  to the open Publications published on HAL : https://universite-paris-sacl
 ay.hal.science/LISN/search/index?q=blot
CATEGORIES:M-E,Thèses et HDR
LOCATION:LISN Site Plaine – Digitéo\, 1 rue René THOM 91190 Gif-sur
 -Yvette\, France
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=1 rue René THOM 91190 Gif-
 sur-Yvette\, France;X-APPLE-RADIUS=100;X-TITLE=LISN Site Plaine – Digit
 éo:geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:STANDARD
DTSTART:20251026T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
END:VTIMEZONE
END:VCALENDAR