Saturday 1st October 1:00-1:40pm
The estimator expressing the estimator as a mean has a lot of data-independentquantities, including the term with the digamma function, inside the log, and in turn inside the sum. Is there a reason not to pull them out?
One could certainly pull them out. I only put them together to make a more compact expression.
For those interested: here's Richard's paper on arXiv:
Copyrights © 2017 All rights reserved