Method, device and storage medium for improving prediction performance of deep learning network

Abstract

The present application relates to a method, device and storage medium for improving the predictive performance of a deep learning network, which belongs to the field of artificial intelligence and computer technology. The method includes: using a preset training set to perform m rounds of iterations on the first neural network model Training; use the pre-set verification set to verify the trained neural network model; determine the corresponding performance indicators based on the scene where the trained neural network model is applied; select and determine from the neural network model obtained by m rounds of iterative training Multiple neural network models whose performance indicators reach the fusion standard; obtain the network parameters of the selected multiple neural network models, and fuse the network parameters of the multiple neural network models; assign the fused network parameters to the second neural network model A neural network model with fused network parameters is obtained; not only does not increase the time for training the model, but also improves the efficiency of network prediction, and can also meet the different needs of application scenarios.

Description

technical field [0001] The present application relates to a method, device and storage medium for improving the prediction performance of a deep learning network, and belongs to the field of artificial intelligence and computer technology. Background technique [0002] Deep learning refers to a collection of algorithms that use various machine learning (Machine Learning) algorithms to solve various problems such as images and texts on a multi-layer neural network. The core of deep learning is feature learning, which aims to obtain hierarchical feature information through layered networks. In order to improve the prediction accuracy of the neural network model, the neural network models in various situations can be fused. [0003] Model fusion includes single model fusion and multi-model fusion. Single model fusion includes multi-layer feature fusion and network snapshot fusion. Multi-layer feature fusion is to fuse features with complementary information in different layer...

AI Technical Summary

Problems solved by technology

[0005] The present application provides a method, device and storage medium for improving the prediction performance of a deep learning network, which can solve the problems of long training time and long prediction time when improving the prediction performance of a deep learning network through single model fusion or multi-model fusion question

Images