An Optimization-Based Approach for Efficient Neural Network Training #16
Abstract Training deep neural networks often suffers from slow convergence and high computational cost due to the inefficiency of stochastic gradient descent (SGD). This paper explores an alternative adaptive learning…