The maximum likelihood estimate

Another empirical probability measure that we can use is the famous maximum likelihood estimate. It is given by :

$$P(s)=\frac{n_s}{N_s}$$

This probability estimate is frequently used because it is simple and gives generally correct results.

This probability estimation has been used by Guyon and Pereira [29] to train a variable length Markov model on strings. It has also been used by Galata, Johnson and Hogg [28] to train a variable length Markov model on the prototypes derived from flow vectors.