ระบบตรวจจับการบุกรุกเครือข่ายสำหรับสำนักหอสมุดมหาวิทยาลัยเชียงใหม่ โดยการใช้ตัวจำแนกข้อมูลนาอีฟเบส์

Abstract

This independent study aims to develop a prototype to predict computer network intruder packages in the Chiang Mai University Library dataset. The experiment applies Naive Bayes classifier on 300,000 records of traffic data by adjusting parameters in order to acquire the best accurate results. Performance measures consist of accuracy, precision, recall, F-measure, and root mean squared error. In addition, we employ KDD Cup Data 1999 which contains 37 categories of network intrusion to discover intrusion types that compromise the Chiang Mai University Library.