second-order methods

Optimization algorithms that use information about the curvature of the loss function, generally providing faster convergence compared to first-order methods that only use gradient information.

5 papers