<ul><li>Researchers have developed an active learning framework for estimating means of multiple groups with unknown data distributions.</li><li>The framework focuses on collecting data fairly and efficiently, especially in dynamically changing environments like online platforms or healthcare trials.</li><li>An algorithm called Variance-UCB is proposed to select groups based on upper confidence bounds on variance estimates, aiming to minimize collective noise in estimators.</li><li>The framework provides efficient bounds for learning from various distributions and improves upon existing regret bounds while offering new results for different objectives and distributions.</li></ul>

An active learning framework for multi-group mean estimation

Discover more