machine learning in intrusion detection systems (ids)

25
Machine Learning in Intrusion Detection Systems (IDS)

Upload: nuala

Post on 21-Jan-2016

72 views

Category:

Documents


0 download

DESCRIPTION

Machine Learning in Intrusion Detection Systems (IDS). 2 papers:. Artificial Intelligence & Intrusion Detection: Current & Future Directions [AIID] J. Frank Applying Genetic Programming to Intrusion Detection [GP] M. Crosbie, G. Spafford. AIID. What is intrusion detection? - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Machine Learning in Intrusion Detection Systems (IDS)

Machine Learning in Intrusion Detection Systems

(IDS)

Page 2: Machine Learning in Intrusion Detection Systems (IDS)

2 papers:

Artificial Intelligence & Intrusion Detection: Current & Future Directions [AIID]– J. Frank

Applying Genetic Programming to Intrusion Detection [GP]– M. Crosbie, G. Spafford

Page 3: Machine Learning in Intrusion Detection Systems (IDS)

AIID

What is intrusion detection? What are the issues in Intrusion Detection?

– Data collection– Data reduction– Behavior Classification– Reporting– Response

Page 4: Machine Learning in Intrusion Detection Systems (IDS)

AIID

AI methods are used to help solve some issues

For data classification:– Classifier systems

• Neural Network

• Decision Tree

• Feature Selection

Page 5: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Data Reduction– Data Filtering– Feature Selection– Data Clustering

Page 6: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Behavior Classification– Expert Systems– Anomaly Detection– Rule-Based Induction

Page 7: Machine Learning in Intrusion Detection Systems (IDS)

AIID

An experiment using Feature Selection– Info. about network connections using a

Network Security Monitor

Page 8: Machine Learning in Intrusion Detection Systems (IDS)

AIID

3 Search algorithms used:– Backward Sequential Search (BSS)– Beam Search (BS)– Random Generation Plus Sequential Selection

(RS)

Page 9: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Algorithm performance

Page 10: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Error Rate Performance (All)

[I, W, T, PS, PD, DS]

[T, PD, DS]Best

Page 11: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Error Rate Performance (SMTP)

[W, T, PS, PD, DS]

Best

Page 12: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Error Rate Performance (Login)

Best[W, T, PS, PD]

[T, PD, DS]RGSS

Page 13: Machine Learning in Intrusion Detection Systems (IDS)

AIID

Error Rate Performance (Shell)

[W, PS, PD, DS]BS & BSS

Best

[W, T, PS, DS] RS

Page 14: Machine Learning in Intrusion Detection Systems (IDS)

GP (Applying Genetic Programming to Intrusion Detection)

An IDS that exploits the learning power of Genetic Programming

Two types of security tools :– Pro-active– Reactive : IDS falls in this catergory

Page 15: Machine Learning in Intrusion Detection Systems (IDS)

GP

Components in an IDS– Anomaly

• May indicate a possible intrusion

– So how do we know for sure? Expert-system• Rule-set = model• Metrics• Comparing metrics & model

But …If a new intrusion scenario arises modifying the

IDS is complicated

Page 16: Machine Learning in Intrusion Detection Systems (IDS)

GP

A finer-grained approach

IDS gets split into multiple Autonomous Agents

Page 17: Machine Learning in Intrusion Detection Systems (IDS)

GP

Page 18: Machine Learning in Intrusion Detection Systems (IDS)

GP

Using GP for learning– Instead of a monolithic static “knowledge base”– The GP paradigm allows evolution of agents

that could be placed in a system to monitor audit data

– GP programs • are in a simple meta-language

• Have primitives that access audit data fields and manipulate them

Page 19: Machine Learning in Intrusion Detection Systems (IDS)

GP

Internal agent architecture

Page 20: Machine Learning in Intrusion Detection Systems (IDS)

GP

Learning by feedback What do the agents monitor?

– Inter-packet timing metrics:

Total # of socket connections, average time between socket connections, minimum time between socket connections, maximum time between socket connections, destination port, source port

– Potential intrusions looked for:Port flooding, port-walking, probing, password cracking

Page 21: Machine Learning in Intrusion Detection Systems (IDS)

GP

Δ = | outcome – suspicion |

Penalty = Δ * ranking /100

Fitness = (100 – Δ) - penalty

Page 22: Machine Learning in Intrusion Detection Systems (IDS)

GP Multiple types:

– Time (long int), port (int), boolean, suspicion (int) Problems with multiple types ADF solution to type safety

– ADF: Automatically Defined Function– To monitor network timing:

avg_interconn_time, min_interconn_time, max_interconn_time

– For port monitoing:src_port, dest_port

– For privileged port checking:is_priv_dest_port, is_priv_src_port

Page 23: Machine Learning in Intrusion Detection Systems (IDS)

GP

Experimental results:

Page 24: Machine Learning in Intrusion Detection Systems (IDS)

That’s it !!!

Page 25: Machine Learning in Intrusion Detection Systems (IDS)

Too old a research idea … did not find any current researches in the same field