Model fitting by least squares |
Nonlinearity arises in two ways: First, modeled data might be a nonlinear function of the model parameters. Second, observed data could contain imperfections that force us to use nonlinear methods of statistical estimation.