In: Statistics and Probability
2. A jar contains 5 balls, 4 of which are blue and one red.
(a) If you draw balls one at a time and replace them, what is the expected draw at which
you will see the red ball?
(b) If you draw balls one at a time, but without replacing them, what is the expected time
to see the red ball?
3. A jar contains 1 red ball and an unknown number of blue balls. You make 20 draws with
replacement from the jar. What is a maximum likelihood estimator of the number of blue
balls in the jar?
2.
(a)
Suppose, random variable X denotes number of draws required to get red ball.
As we draw with replacement, X can take any value from 1 to infinity.
Now, probability to be selected in i th draw
[ Same as geometric distribution, as getting red ball can be termed as success]
Subtracting second from first we get,
[ Infinite GP series]
So, expected number of draws to see red ball is 5.
(b)
Suppose, random variable Y denotes number of draws required to get red ball.
As we draw without replacement, Y can take any value from 1 to 5.
P(Y=y) depends on each draw of 1,2,...,y. As ball is not replaced, a ball other than red must be drawn out of remaining in each earlier steps.
So, expected number of draws to see red ball is 3.
2.
Suppose there are n blue balls in the jar.
So, in total n+1 balls.
Balls are drawn with replacement.
So, as of 1(a) we can calculate and conclude that expected number of draw is n+1 to get the red ball.
But, it is given that n+1=20.
So, maximum likelihood estimator of number of blue balls in the jar,