Abstract: Traditional uncertainty optimization methods employ a two-stage paradigm that separates uncertainty estimation from decision-making, often yielding suboptimal solutions. To address this ...