If the model depends linearly on the unknown parameters
to be estimated from the data, it is called a "linear model” and
also the regression is “linear”. But, often a good agreement
between model and available measurements requires a nonlinear
dependence on the sought parameters, i.e. non-linear
regression occurs and, consequently, non-linear least squares
arise. From the point of solution of the equation system (3), a
linear model involves the existence of a closed form solution.
Otherwise, a non-linear model implies that there is no closedform
solution. Instead, numerical algorithms are used to find
the value of the parameters which minimize the objective.
Most algorithms involve choosing initial values for the
parameters. Then, the parameters are refined iteratively, that
is, the values are obtained by successive approximation.