Extended formulation for unscented transform and its application as Monte Carlo alternative
José Edil Guimarães de Medeiros, S.A.P. Haddad, L.R.A.X. de Menezes · Electronics Letters · 2016
A novel approach for the unscented transform calculation is detailed, showing a technique that can be used to obtain smoother estimates for probability functions based on Monte Carlo technique substitution. The method is computationally efficient for statistical moment estimation of nonlinear transformations and does not require full knowledge of the probability distribution of the random variables, being completely defined by its moments.