On sparsity-exploiting memory-efficient trust-region regularized nonlinear least squares algorithms for neural-network learning.

Eiji Mizutani, James Weldon Demmel

Browse the full IJCNN paper archive.