Comment Comment by reader5000 Great. Explain why training converges in subexponential time despite the nonconvex cost function and does not overfit the data. parent reader5000 9 years ago
