gradient descent

Gradient descent is an optimization algorithm used to minimize the loss function of machine learning models by iteratively adjusting the model parameters in the direction of the steepest descent, based on the computed gradients.

51 papers