首页|A Metaheuristic Optimization Approach-Based Anomaly Detection With Lasso Regularization
A Metaheuristic Optimization Approach-Based Anomaly Detection With Lasso Regularization
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Igi Global
The paper suggests a new intrusion detection method that is applied using an improved Lasso Regularization Metaheuristic Optimization. A comparison with other metaheuristic algorithms used to test the current method alongside the associated works is part of the proposed analysis herein. Because of the high processing capacity, network traffic in the intrusion detection system (IDS) has erratic behavior. The size of the device increases; the vast number of features must then be explored. However, the undesirable features and (or) any noisy data have a significant effect on the performance of the IDSs. Lasso regression implements L1 regularization, applying a penalty proportional to the absolute value the coefficient magnitude. Sparse formulas with few coefficients can result in this form of regularization; certain coefficients can become negative and be removed from the model. The algorithm for particle swarm optimization (PSO) is added to the selective features that increased the IDS detection rate and accuracy.
NETL
NSTL
Igi Global
