Biometrics - Parameter Estimation
The estimated amount required parameter estimation of unknown parameters and a certain degree of confidence
Estimation Methods: estimating a point estimate value; using interval estimates may interval and the estimated value of the likelihood value.
Evaluation criteria of estimates:
No bias is to estimate the amount (not necessarily the sample mean) Sampling mathematical expectation distribution and population parameters such as the true value.
Sometimes several sets of data is valid is unbiased, but this time is the minimum number of effective variance.
Consistency means that the sample is large, it is estimated more accurate.
Existing methods are by these three evaluation criteria.