- noun the calculation of parameter values which fit the observed data best
- (written as Maximum Likelihood Estimation)Statistical estimation using the criterion of what is most likely to be the case. A method of estimation that specifies the probability of the observed set of data and finds the parameter values that maximize that probability. For example, if the sampled variable in which you are interested is normally distributed, the mean is the maximum likelihood estimator of the population mean, and the variance is an approximation to the maximum likelihood estimator of the population. The idea was that of the English statistician Sir Ronald Fisher (1890-1962) (see Fisher, 1922).