THE APPLICATION OF STATISTICS
INTO THE INTELLIGENT DETECTION
OF POTENTIAL CYBER THREATS
•Authors: Jie Zhao*, Trey Yeager*
•Date: February 8, 2015
•Class Project: Problem Introduction
Problem
description & motivation
•
Cybersecurity is the protection of on-line data and services through the prevention of cyber attacks.
•
Failing to prevent a cyber attack can result in the loss of important data or services. In today's world there is a lot of damage that can be caused by successful hacker attacks. •
By analyzing the current datasets it may be possible to derive a model, using statistics, that can automate the detection of threats on a network.
•
One of the biggest implications of this study would be the improved security of networks. This means better protection from cyber terrorists, better protection of an individual's on-line data, and better protection of different industries from potential threats. •
The statistics will be used to create a model that determines if a network is currently under attack. In addition, the statistics will be used to determine if there is any interactions between predictors like physical devices (Router, Gateway, server, workstation) and type of attacks.
•
The statistical model will be a prediction model that helps determine if a network is currently under attack.
•
Using a statistics model to expose a hacker's attack will allow us to better understand network information and how it applies to cyber security.
Problem
statement
• CAIDA (Center for Applied Internet Data Analysis) has been collecting internet data from different networks that have been subjected to hacker's attacks. We will use this data to create a model that can determine if the using network is currently under attack. • Previously without any statistics-based detection, it has been easier for hackers to bypass network defenses. Because the hackers always routinely bleach the defense with
