(Retracted) Research on computer network security evaluation based on image recognition and neural network

Abstract

In today’s social development, the level of informatization has been greatly improved. Computer networks are now used in all aspects of society, especially with the popularity of mobile phones and computers. Basically, a mobile phone is a computer that stores all the information of a person, so it is indispensable for carrying out research on computer network security evaluation. The purpose of this article is to examine the growing state of information security threats in computer networks. Neural network technology, which is considered secure, has high flexibility, fast access to information, strong logic, and low failure. Toward creating a low-cost network topology, we used a low-cost network risk assessment method to assess the security of computer networks based on neural networks, and we evaluated existing risks that are appropriate for future computer security research. The results of the study show that neural networks using the Bayesian network method have a significant impact on network security. Considering the problems that can occur with information security risks, computer security systems based on neural networks have a 30% reduction in information error rate compared with traditional methods, and computer network systems based on neural networks have better learning performance to safeguard network security. Improving the security of information requires further research and evaluation of computer network security based on neural networks.