Enhanced battery life prediction with reduced data demand via semi-supervised representation learning

A semi-supervised learning method is proposed to integrate both labelled and unlabelled data during model training, resulting in enhanced RUL prediction performance compared to supervised models trained solely on labelled data.