An Improved Intrusion Detection Algorithm Based on GA and SVM

In the era of big data, with the increasing number of audit data features, human-centered smart intrusion detection system performance is decreasing in training time and classification accuracy, and many support vector machine (SVM)-based intrusion detection algorithms have been widely used to ident...

Full description

Saved in:
Bibliographic Details
Published inIEEE access Vol. 6; pp. 13624 - 13631
Main Authors Tao, Peiying, Sun, Zhe, Sun, Zhixin
Format Journal Article
LanguageEnglish
Published Piscataway IEEE 01.01.2018
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Biological cells
Classification
Crossovers
Feature extraction
Feature selection
Fitness
Genetic algorithm
Genetic algorithms
Intrusion detection
Intrusion detection systems
Optimization
Parameters
Sociology
Statistics
support vector machine
Support vector machines
Online AccessGet full text

Cover

More Information
Summary:In the era of big data, with the increasing number of audit data features, human-centered smart intrusion detection system performance is decreasing in training time and classification accuracy, and many support vector machine (SVM)-based intrusion detection algorithms have been widely used to identify an intrusion quickly and accurately. This paper proposes the FWP-SVM-genetic algorithm (GA) (feature selection, weight, and parameter optimization of support vector machine based on the genetic algorithm) based on the characteristics of the GA and the SVM algorithm. The algorithm first optimizes the crossover probability and mutation probability of GA according to the population evolution algebra and fitness value; then, it subsequently uses a feature selection method based on the genetic algorithm with an innovation in the fitness function that decreases the SVM error rate and increases the true positive rate. Finally, according to the optimal feature subset, the feature weights and parameters of SVM are simultaneously optimized. The simulation results show that the algorithm accelerates the algorithm convergence, increases the true positive rate, decreases the error rate, and shortens the classification time. Compared with other SVM-based intrusion detection algorithms, the detection rate is higher and the false positive and false negative rates are lower.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2018.2810198