Approximation-based iterative learning control for uncertain It o ^ stochastic nonlinear systems with arbitrary initial shifts | Synapse