factorization methods for discrete sequential esti future state depends only on a finite number of past states: \[ P(X_t | X_{1:t-1}) \approx P(X_t | X_{t-k:t-1}) \] for some order \( k \). First-order Markov Chain: \[ P(X_1, ..., X_T) = P(X_1) \prod_{t=2 T Toby Jacobs Oct 17, 2025
