Reverse Back Propagation to Make Full Use of Derivative
This is an incremental improvement for neural network training efficiency without extra inference costs.
The paper tackles the problem of optimizing input loss at the neural network input end by reversing the back-propagation process, resulting in adaptation to a larger learning rate range and better learning than vanilla back-propagation on MNIST, CIFAR10, and CIFAR100 datasets.
The development of the back-propagation algorithm represents a landmark in neural networks. We provide an approach that conducts the back-propagation again to reverse the traditional back-propagation process to optimize the input loss at the input end of a neural network for better effects without extra costs during the inference time. Then we further analyzed its principles and advantages and disadvantages, reformulated the weight initialization strategy for our method. And experiments on MNIST, CIFAR10, and CIFAR100 convinced our approaches could adapt to a larger range of learning rate and learn better than vanilla back-propagation.