An Iterative Nonlinear Regression for Polarisation/depolarisation Currents

To define the relationship between the response – a random or dependent variable y and a predetermined or independent variable x, a straight line or concave respectively convex functions affected by random errors must be used. A locally weighted low-degree polynomial regression i.e. linear is fit to several subsets of data using linear least square fitting and a first-degree polynomial or the exponential curve. PDC RVM Nova software provides a toolbox to load the dataset, to filter the data and to estimate the parameters of each subset.