Page 85 - Probability and Statistical Inference
P. 85

62    1. Notions of Probability

                                 θ and β



                                                     +
                                 where θ ∈ ℜ and β ∈ℜ . Here, θ and β are referred to as parameters. In the
                                 statistical literature, this is known as the Laplace or double exponential distri-
                                 bution. The Figure 1.8.2 gives a plot of this pdf when θ = 0 and β = 1, 2. It is
                                 implicit that x, f(x) are respectively plotted on the horizontal and vertical axes.
                                       (i)  Show that the pdf f(x) is symmetric about x = θ
                                       (ii)  Derive the expression of the df explicitly;
                                       (iii)  Show that P{| X – θ| > a + b | |X – θ| > a} does not depend on
                                            a, for any two positive numbers a and b;
                                       (iv)  Let θ = 0 and β = 1. Obtain an expression for the right hand tail
                                            area probability, P(X > a), a > 0. Then compare tail area prob-
                                            ability with that of the Cauchy distribution from (1.7.31). Which
                                            distribution has “heavier” tails? Would this answer change if θ =
                                            0 and β = 2 instead? {Hint: Recall the types of discussions we
                                            had around the Table 1.7.1.}
                                    1.7.14 (Exercise 1.6.7 Continued) Consider a random variable X having the
                                 pdf f(x) = c(x – 2x  + x )I(0 < x < 1) where c is a positive number and I(.) is the
                                                2
                                                   3
                                 indicator function. Find the value of c. {Hint: In order to find c, should you
                                 match this f(x) up against the Beta density?}
                                    1.7.15 Show that
                                 0) given by (1.7.27) is a bona-fide pdf with µ ∈ ℜ, σ ∈ ℜ . Also verify that
                                                                                 +
                                                                                 +
                                       (i)  P(X ≤ x) = Φ({log(x) – µ}/σ), for all x ∈ ℜ ;
                                       (ii)  P{log(X) ≤ x} = Φ({x – µ}/σ), for all x ∈ ℜ.
                                    {Hint: Show that (a) f(x) is always positive and (b) evaluate    by
                                 making the substitution u = log(x).}
                                    1.7.16 The neighborhood fishermen know that the change of depth (X,
                                 measured in feet) at a popular location in a lake from one day to the next is a
                                 random variable with the pdf given below:


                                 where c is a positive number. Then,
                                       (i)  find the value of the constant c;
                                       (ii)  determine the expression of the df;
                                       (iii)  find the median of this distribution.
                                    1.7.17 The neighborhood fishermen know that the change of depth (X,
                                 measured in feet) at a popular location in a lake from one day to the next
   80   81   82   83   84   85   86   87   88   89   90