本文研究了正则化格式下随机梯度下降法的收敛速度.利用线性迭代的方法,并通过参数选择,得到了随机梯度下降法的收敛速度.%This paper investigates the learning rates of the stochastic gradient descent methods in regularization schemes.By the iteration method,we obtain the convergence rate of the algorithm with suitable choice of the parameter.
展开▼