Normal Distribution - Parameter Estimation |
|
The normal probability density function is , -¥<x<¥ where m is the distribution mean, and If no censoring is involved, the distribution mean is estimated from the expression where n is the sample size. If no censoring is involved, the distribution standard deviation is estimated from the expression However, when censored data are involved, parameter estimation becomes complicated. Three popular methods for parameter estimation for the normal distribution when censored data are encountered are After distribution parameters have been estimated, reliability estimations and predictions are used for evaluations. The maximum likelihood estimation section explains how this can be done manually, but because of the complexity of the calculations, manual methods are not recommended. The predicting module explains how to estimate reliability using The Reliability and Maintenance Analyst software package. |