Probabilistic time series forecasting with compositional bayesian neural networks
AutoBNN combines the interpretability of conventional probabilistic approaches with the scalability and adaptability of neural networks for constructing subtle time ...
