Log Normal distribution (continued)
Can have 2 or 3 parameters
- Parameterized by the moments of Y = ln X
- Relationship between moments of X and Y
- E[X] = exp { E[Y] + var[Y]/2 }
- var[X] = E[X]2 exp { var[Y] - 1 }, or
- Cv = { exp (var[Y]) - 1 }1/2
- Third parameter is offset { Y = ln ( X - t ) }
- Coefficient of skewness is related to the mean and variance of X
- Cs = 3Cv + Cv3 where Cv is of X