I noticed that after calling the score function of "HawkesExpKern" model the returned value did not coincide with the log-likelihood value of a Hawkes Process over the same timestamps and parameters. Is any way to transform the score() function value in order to get the true log-likelihood?
Hello,
I noticed that after calling the score function of "HawkesExpKern" model the returned value did not coincide with the log-likelihood value of a Hawkes Process over the same timestamps and parameters. Is any way to transform the score() function value in order to get the true log-likelihood?
Thanks in advance.