Discrete Probability Models with Modified Zeros
James J. Higgins, Chris P. Tsokos · IEEE Transactions on Reliability · 1978
Two types of discrete probability models are investigated. The models arise from the Poisson process by modifying in certain ways the probability of observing a zero. The first type is a composite of two discrete pdf's to allow for adjustment in the probability of observing no occurrences of the event of interest. The second is a pure birth process in which the transition intensity changes after the first event of interest has occurred. Procedures for estimating the parameters of these models are given. The models fit well the actual data selected from reliability and queuing applications.