Predicting Students' Behavioral Patterns in University Networks for Efficient Bandwidth Allocation: A Hybrid Data Mining Method (Application Paper)

摘要: The effective bandwidth management in multi-service computer networks such as university has become a challenge recent years. growth of internet traffic and limitation resources persuade the information technology (IT) managers to focus on allocation policies. One important issues discussed this domain is how assign fairly based priority levels. In paper, focusing "priority-based allocation", hybrid data mining method developed manage limited network more effectively. This composed two main steps uses clustering classification techniques. purpose detect, analyze predict students' behavioral patterns identify factors that affect their tendency using internet. proposed applied real university. results indicate "degree level" "age" are most influence use would be also useful for prediction purposes. It helps IT new student's given his/her characteristics. By analyzing results, can make better decisions optimize resources.