一种改进的贝叶斯网算法在入侵检测中的应用研究

摘要

This improved Bayesian network algorithm uses rough set theory for data dimension reduction process to reduce the computational cost of Bayesian network learning. Using the theory of mutual information, a Bayesian network structure learning algorithm based on mutual information is proposed. Experiments show that this algorithm can improve the detection precision and the detection efficiency.%这种改进的贝叶斯网算法采用粗糙集理论对数据集进行降维处理，减少贝叶斯网学习的计算成本；利用互信息理论，提出一种基于互信息的贝叶斯网结构学习算法。实验表明这种算法，即提高了检测精度又提高了检测效率。