Multi-target semi-supervised learning with application to small area estimation

Abstract

This invited talk extends semi-supervised learning to settings with several quantities to estimate across subpopulations that have few labelled observations. It discusses estimators designed to improve stability and efficiency, and shows how small area estimation fits into this framework.

Date
Jun 23, 2026 9:00 AM — 9:30 AM
Location
Thessaloniki, Greece

Invited talk in the session “Contributions to Understandable Machine Learning”.

Katarzyna Reluga
Katarzyna Reluga
Assistant Professor