1 paper
Yuka Hashimoto, Tomoharu Iwata
We propose deep Koopman-layered models with learnable parameters in the form of Toeplitz matrices for analyzing the transition of the dynamics of time-series data. The proposed mod…