Development of an Approach to Confidential Learning with Errors in the Design of Neural Networks

In this paper, the presented method, applied to transpose densified weight matrices, involves tenth machine learning when moving from feedforward to backpropagation. The algorithm is based on the diagonal matrix construction method. The process of manufacturing change will significantly increase the number of operations using technology, especially in the ten linear transformations. The number of multiplications required for linear transformations depends on the dimensionality of the input and output data, but these differences are taken into account during the training process, which includes both forward and back propagation. The proposed method leads to improved training efficiency and computational efficiency in contextual machine learning.