NAVER LABS Europe

20th - 22nd September 2017

On Thursday September 21st, invited speaker Tomi Silander will be presenting "Hyperparameter sensitivity revisited". 

Abstract: The BDeu scoring criterion for learning Bayesian network structures is known to be very sensitive to the equivalent sample size hyper-parameter. Recently some authors have suggested alternative Bayesian scoring criteria that appear to behave better than BDeu. So is the problem solved? We will review the problem and suggested solutions and present empirical assessment of the current situation.