Network model construction method and system based on distributed algorithm

Abstract

The invention provides a network model construction method and system based on a distributed algorithm, and the method comprises the steps: obtaining a target data set inputted by a user in real time, and carrying out the full scanning of the target data set, so as to detect a plurality of data types correspondingly contained in the target data set in real time; matching a data subset corresponding to each data type in the target data set in real time, and generating a plurality of corresponding data families in real time according to each data type and the corresponding data subset through a preset distributed algorithm; and detecting a plurality of training nodes correspondingly contained in the preset neural network in real time, and performing iterative training on each training node through each data family to train a corresponding target model in real time. According to the method, the training of various models can be correspondingly completed under the condition of saving the computing power, meanwhile, the training time can be shortened, and the working efficiency is correspondingly improved.

Images