NeurIPS 2020

Deep reconstruction of strange attractors from time series


Meta Review

This paper uses a novel regularizer based on false neighbors to find latent delay embedding. There were intense discussions among the reviewers on sec 2. I tend to agree with R1 who puts more emphasis on the innovation of the approach. However, I found the important details of the method is hidden in the supplement. I suggest switching parts of sec 2 with the details about the method from the supplementary text.