A Distributional Optimisation Perspective on Combining Models in Deep Learning

Congye Wang, Yan Lin, Zheyang Shen +2

Combining predictions from different models can improve performance at machine learning tasks, but the training of the individual models and the rule used to combine them are typic…