The dice factory


On: Bayesian analysis, Murphy Ch. 5

Question: Joseph is a recently-hired data scientist who has been asked to evaluate the fairness of dice from a recently-opened factory. His client requested that Joseph roll the dice in twos but that he not roll the dice "too many times" to avoid wear on the dice, expressing confidence that the dice were "okay". To incorporate his client's belief that all sides have equal probability of landing face up, Joseph decides on a Bayesian analysis for credible intervals and sets out to use appropriate hyperparameters. What distribution is mathematically convenient in describing his client's belief?

A. the Gaussian distribution
B. the Bernoulli distribution
C. the multinomial distribution
D. the Dirichlet distribution

Course Home | Instructor | Schedule | Readings