A Distance for HMMs based on Aggregated Wasserstein Metric and State Registration

We propose a framework, named Aggregated Wasserstein, for computing a dissimilarity measure or distance between two Hidden Markov Models with state conditional distributions being Gaussian. For such H