Statistical and machine learning approaches for the minimization of trigger errors in parametric earthquake catastrophe bonds