Dynamic sample size selection based quasi-Newton training for highly nonlinear function approximation using multilayer neural networks.

Hiroshi Ninomiya

Browse the full IJCNN paper archive.