New method revolutionizes system identification for improved accuracy and stability.
A new method called compensated total least squares is used to estimate models with linear or nonlinear relationships between variables. This method corrects for biases and is reliable and accurate. The estimates obtained are proven to be consistently accurate through simulations.