Consider a string of Bernoulli trials with probability of success equal to p and let Y1 be the number of trials untial the first success and Y2 be the number of trials untial the second success. Y1 is therefore a geometric random variable and Y2 is a negative binomial random variable. Note that once you know on which trial the first success occurred, the number of additional trials (y2-y1) to get the second success is also a Geometric random varriable. Develop an expression for the condional probability that Y2=y2 given Y1=y1. Use this to derive the joint probability function of Y1 and Y2. What is the covariance of Y1 and Y2