Can I feed **_unique temporal walks_** (i.e., random walks forward in the time) as **_ngrams_** to calculate the optimal k-th order of the underlying network?
Can I feed unique temporal walks (i.e., random walks forward in the time) as ngrams to calculate the optimal k-th order of the underlying network?