ON THE PROBABILITY DISTRIBUTION OF THE CLASSICAL GRAS IMPLICATION INDEX BETWEEN TWO BINARY RANDOM VARIABLES
Pablo Gregori Huerta and Raphael Couturier and Rubén Pazmiñno Maji
In this contribution we study the behavior of the classical Gras implication index as a random variable, when applied to a couple of Bernoulli variables (x;y), independent or not. We also show the efect of the conditional probability Py|x on its probability distribution, and specially on its mean value and quartiles.
Keywords: binomial model, classical implication index, conditional probability, multinomial model.