Serving the Quantitative Finance Community
ML estimation with the t-distribution assumption
for data analysis based in time series, anyone knows how to compute the ML estimator in t-innovation, known as its degree of freedom?the key point however might be to look for a numerical technique to differentiate gamma functiontks a lot!
Last edited by GP03 on February 8th, 2009, 11:00 pm, edited 1 time in total.
ML estimation with the t-distribution assumption
I think that it usually done numerically rather than in closed form.But you could use the digamma function if you have access to that [see Abramowitz + Stegun]Kyriakos