摘要
随着大数据环境下网络安全问题的日益复杂,传统的漏洞检测方法面临过拟合与高资源消耗问题。提出一种基于DNN的漏洞检测与防护技术,通过自适应正则化和稀疏化技术,优化了检测模型,提升了检测准确性和计算效率。实验结果表明,优化后的模型在准确率、召回率、检测速度及资源消耗方面表现优异,且防护策略通过强化学习显著提升了防护能力和响应速度。
With the increasing complexity of network security issues in the big data environment,traditional vulnerability detection methods face problems of overfitting and high resource consumption.This paper proposes a vulnerability detection and protection technology based on DNN,which optimizes the detection model through adaptive regularization and sparsity techniques,and improves detection accuracy and computational efficiency.The experimental results show that the optimized model performs well in accuracy,recall,detection speed,and resource consumption,and the protection strategy significantly improves the protection ability and response speed through reinforcement learning.
作者
梁仲峰
LIANG Zhongfeng(Guangxi Power Grid Co.,Ltd.,Nanning,Guangxi 530000,China)
出处
《自动化应用》
2025年第18期267-269,共3页
Automation Application
关键词
大数据
安全漏洞
深度神经网络
优化算法
防护策略
big data
security vulnerabilities
deep neural network
optimization algorithm
protection strategy
