The researchers developed a new method for estimating parameters in a mathematical model that combines polynomials and sinusoidal functions. This method, called weighted least squares, is more robust against errors than the traditional least squares method, especially when dealing with outliers in the data. The weighted least squares method is easier to use than other robust methods and performs well in simulations. Choosing the right weight function is important for the method's success. The researchers tested their method on real data and found it to be effective.