A Python toolbox/library for reality-centric machine/deep learning and data mining on partially-observed time series with PyTorch, including SOTA neural network models for science tasks of imputation, classification, clustering, forecasting & anomaly detection on incomplete (irregularly-sampled) multivariate time series with NaN missing values/data
Why do you think that https://github.com/datamllab/tods is a good alternative to PyPOTS