Theoretical notions of statistical estimation

This article will address traditional assessment methods, such as maximum likelihood, useful when it is known. Conversely, it is not known, we can use nonparametric methods exploiting specific property that require the involvement of a distribution functions. Models with discrete variables and partially observed models are usually estimated by maximum likelihood method. We will address some models presented in the first section of this chapter and will use their traditional presentation as a model of parametric indices. We will address the theory of regression observation asymptotically unbiased estimator using positive and analyzing their effectiveness. We will further analyze the types of errors that are generated from regression and heteroscedastic.