An Improved Approach for Cluster Newton Method in Parameter Identification for Pharmacokinetics

This paper proposes an improved scheme for the original cluster Newton method to contemporaneously find multiple solutions for inverse parameter identification in pharmacokinetics by applying Tikhonov regularization for fitting a super plane in the CN method. The numerical experiments of the proposed approach have proven that when using these proposed approaches, we can (i) save iterations; (ii) save computation time; and (iii) the cluster of points moves more stably toward the manifold of solutions.