I recently published a preprint on ensemble learning with calibrated losses. For further details, please refer to the arXiv and GitHub links.