This paper focuses on enhancing the performance of Intrusion Detection Systems (IDS) within the domain of network security through data preprocessing and algorithm optimization, addressing the increasingly complex landscape of network threats. This study utilizes the NSL-KDD dataset, addressing high dimensionality with Principal Component Analysis (PCA) and handling outliers via RobustScaler. Logistic regression and decision tree models were used for classification, with random oversampling applied to mitigate class imbalance. The main innovation of this study is the integration of PCA and decision tree models, which improves the detection accuracy in high-dimensional data, as well as the effective use of random oversampling to address class imbalance issues. The experimental results showed that the decision tree model surpassed logistic regression, achieving a prediction accuracy of 99.63% compared to 91%, confirming the superiority of nonlinear classification models for this datasets. Moreover, the balanced accuracy of the model improved after applying random oversampling, proving the efficacy of this method in addressing minor class imbalance. This study demonstrates that a combination of PCA and random oversampling techniques can significantly enhance IDS performance in detecting complex network threats. The results also indicate that nonlinear models such as decision trees are better suited for handling high-dimensional and imbalanced datasets, providing important insights for improving IDS performance. Future research may consider integrating more sophisticated deep learning methods and leveraging real-time big data analytics to further improve IDS generalization and responsiveness to emerging threats.

Tengyao Wang. Research on Intrusion Detection Classification Methods Based on PCA-Decision Tree. Academic Journal of Computing & Information Science (2024), Vol. 7, Issue 12: 44-50. https://doi.org/10.25236/AJCIS.2024.071206.

[1] Zhou Qi. Research on Intrusion Detection Model Based on Feature Selection and Machine Learning Algorithms [D]. Gannan Normal University, 2023.

[2] Shang Jiaqi. Research on Intrusion Detection Models for Imbalanced Data [D]. Liaoning: Liaoning Technical University, 2023.

[3] Dong Yifan. Study on the Robustness of Intrusion Detection Methods in Adversarial Environments [D]. National University of Defense Technology, 2018.

[4] Li Yinghao. Research on PCA-based Feature Subset Selection and Network Intrusion Detection Algorithm [D]. Tianjin: Tianjin University of Technology, 2021.

[5] Ma Zehui. Research on Webshell Detection Method Based on Logistic Regression Algorithm [J]. Information Security Research, 2019, 5(4):298-302.

[6] Xu Ying, Qu Danqiu. Design and Implementation of a Computer Network Intrusion Detection System Based on Decision Tree Classification Algorithm [J]. Information Recording Materials, 2024, 25(6): 137-139. 

[7] Jian Shijie. Research on Intrusion Detection Method Based on Oversampling and Traffic Data Dimensionality Reduction [D]. University of Chinese Academy of Sciences, 2021.