1 paper
Mojtaba Moattari, Mohammad Hassan Moradi, Reza Boostani
Gradient Descent (GD) approximators often fail in the solution space with multiple scales of convexities, i.e., in subspace learning and neural network scenarios. To handle that, o…