分析駭客入侵資料程度之混合模型

Analyzing the Degree of Hacker Intrusion Data with Hybrid Model

陣博德(Po-Te Chen)

指導教授 ： 黃俊哲

國立暨南國際大學/管理學院/資訊管理學系/碩士(2018年)

摘要

如何對駭客進行防範，找出關鍵的目標，並實施有效的決策，對於資訊安全是最重要的問題，本研究提出一個混合模型的分析方法，從入侵偵測模型規則中找出規則，首先透過邏輯回歸進行預測分析找出遭到誤判的目標對象，這些對象在預測中應該安全，但實際遭到攻擊，因此對於遭到誤判的目標對象實施決策是最有效益的，接下來使用粗糙集進行約簡與核，發現變項屬性中隱藏的規則，在對遭到誤判的目標對象進行規則比較，找出遭到誤判的目標對象被攻擊的關鍵因素，最後提出屬性改善表，提供決策者去進行決策。本研究透過資訊安全中入侵偵測的案例，使用本研究所提出的方法去進行分析，從規則數據中產生出屬性改善表，實施有效的決策。

關鍵字

粗糙集 ； 邏輯回歸 ； 資訊安全 ； 入侵偵測 ； 混合模型 ； 誤判

並列摘要

How to prevent hackers, identify key targets and implement effective decision- making is the most important issue for information security. This study proposes a hybrid model analysis method to find the rules from the intrusion detection model rules. First, predictive analysis through logistic regression to find out the target object that was False Positives, these objects should be safe in the prediction, but they are actually attacked. Therefore, it is most effective to implement decision-making on the target object that has been False Positives. Next, use the rough set for Reduct and core, find the hidden rules in the variable properties, then compare the rules of the target object that was False Positives, Identify the key factors for the target being attacked. Finally, the attribute improvement table is proposed to provide decision makers to make decisions. The intrusion detection case in information security is studied to validate superiority of the proposed solution approach.

並列關鍵字

Rough sets ； Logistic regression ； Information security ； Intrusion detection ； Hybrid Model ； False Positives

參考文獻

一、英文部分

Denning, D. E. (1987). An intrusion-detection model. IEEE Transactions on software engineering, (2), 222-232.

Kemmerer, R. A., & Vigna, G. (2002). Intrusion detection: a brief history and overview. Computer, 35(4), supl27-supl30.

Tajbakhsh, A., Rahmati, M., & Mirzaei, A. (2009). Intrusion detection using fuzzy association rules. Applied Soft Computing, 9(2), 462-469.

Thaseen, I. S., & Kumar, C. A. (2017). Intrusion detection model using fusion of chi-square feature selection and multi class SVM. Journal of King Saud University-Computer and Information Sciences, 29(4), 462-472.