Neural network-based finite-horizon approximately optimal control of uncertain affine nonlinear continuous-time systems | IEEE Conference Publication | IEEE Xplore