First I would like to say I read AIMA 2nd edition for fun and it was fantastic.
I have been studying the hybrid Bayesian network for an application and am wearing out
pages 501 and 502 which talk about combining a discrete true/false parent (Subsidy) and
a continuous parent (Harvest) into a linear Gaussian distribution.
I am simply not seeing where the variables required as parameters for the Gaussian are
coming from. It seems that the h variable is the Harvest, and Subsidy would be some
The extra variables are:
a-true, b-true, stddev-true
a-false, b-false, stddev-false
Would someone take pity on me? :)