Click for printer friendely version of this HowTo Overview Maximum Likelihood is a method for estimating parameters for distributions. For example, if you have a set of independent3.4 data points, X, where X is a vector of points that have been collected from the same normal distribution where $\mu$ and $\sigma^2$ are unknown, then maximum likelihood would derive $\hat{\mu}$ and $\hat{\sigma}^2$, estimates for $\mu$ and $\sigma^2$ respectively. The maximum likelihood estimates (MLEs) are such that they maximize the probability or likelihood of the data, X.

>