Multilayer perceptron algorithms for cyberattack detection

A Multilayer Perceptron (MLP) is a machine learning algorithm capable of classifying large amounts of data and finding patterns in complex datasets. In this paper the MLP algorithm is used to perform intrusion detection based on the Knowledge Discovery and Datamining (KDD) dataset. The results show...

Full description

Saved in:
Bibliographic Details
Main Authors: Palenzuela, Francisco, Shaffer, Melissa, Ennis, Matthew, Gorski, Jeffrey, McGrew, Derek, Yowler, Daniel, White, Daniel, Holbrook, Logan, Chris Yakopcic, Taha, Tarek M.
Format: Conference Proceeding
Language:English
Subjects:
Biological neural networks
Classification algorithms
Computer crime
cybersecurity
Intrusion detection
Multilayer perceptrons
neuromorphic
Topology
Training
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A Multilayer Perceptron (MLP) is a machine learning algorithm capable of classifying large amounts of data and finding patterns in complex datasets. In this paper the MLP algorithm is used to perform intrusion detection based on the Knowledge Discovery and Datamining (KDD) dataset. The results show error minimization during training, as well as classification accuracy for a number of different perceptron topologies. These topologies differ in their number of hidden layers, and in the number of neurons within each hidden layer. The system presented is capable of performing intrusion detection at about 99.99% accuracy after an output bias is applied (with a false positive rate of about 10%).
ISSN:2379-2027
DOI:10.1109/NAECON.2016.7856806