Header menu link for other important links
X
Bias in the maximum likelihood estimation of parameters of nonlinear regression models
B. Mahaboob, , C. Narayana, M. Sivaiah
Published in International Journal of Scientific and Technology Research
2019
Volume: 8
   
Issue: 11
Pages: 1252 - 1255
Abstract
Nonlinear model building has become an important tool in Predictive Analysis and Forecasting Theory. MLE is a phenomenon in which one can obtain unknown coefficients of a distribution by optimizing a likelihood function. Maximum li kelihood estimate is the vector in parameter space which optimizes the likelihood function. This research article throws a light on the BIAS in the MLE of unknown coefficients of statistical models which are not linear. In addition to this a test for the linearity of regression has been proposed. If the ML function possesses derivatives one can apply first derivative test to obtain optimum values. But in some situations the equations of first degree of ML function are to be solved in explicit manner. For example in linear statistical model OLS estimator optimize the ML function. In vast number of cases advanced numerical techniques should be implemented in order to get ML function. As the application of ML technique is both flexible and intuiti ve this technique has become an indispensable tool in statistical inference. ©IJSTR 2019.
About the journal
JournalInternational Journal of Scientific and Technology Research
PublisherInternational Journal of Scientific and Technology Research
ISSN22778616