Two very commonly employed distributions that merely satisfy the above-mentioned
mean and the standard deviation criteria (Equations (3.47) and (3.48)) only are the normal
and the lognormal distributions. However, in the case of a given variable, if the analyst is
forced to select a probability distribution that would represent the random variation of
that variable more accurately, then in addition to the mean and the standard deviation
estimates, one could also compute the coefficients of skewness and kurtosis (flatness)
computed from the sample data (Harr, 1977). It must be noted that the coefficients of
skewness and kurtosis for the population can be related to the third and the fourth
moments of area of the probability distribution about the mean, respectively.
Two very commonly employed distributions that merely satisfy the above-mentioned
mean and the standard deviation criteria (Equations (3.47) and (3.48)) only are the normal
and the lognormal distributions. However, in the case of a given variable, if the analyst is
forced to select a probability distribution that would represent the random variation of
that variable more accurately, then in addition to the mean and the standard deviation
estimates, one could also compute the coefficients of skewness and kurtosis (flatness)
computed from the sample data (Harr, 1977). It must be noted that the coefficients of
skewness and kurtosis for the population can be related to the third and the fourth
moments of area of the probability distribution about the mean, respectively.
การแปล กรุณารอสักครู่..
