PCAmix-based One Class Classification Ensemble Learning for Network Attack Detection
Shaohui Mo, Gulanbaier Tuerhong, Mairidan Wushouer, Tuergen Yibulayin · 2021 International Conference on Computer Information Science and Artificial Intelligence (CISAI) · 2021
In the real network environment, most data is high-dimensional, mixture and unbalanced. In order to process these data, Principal Component Analysis mix (PCAmix) method is used to reduce the dimension of the data, combining the advantages of One Class Classification (OCC) algorithms, Multi-Layer Perceptron (MLP) algorithm and Stacking Ensemble Learning Model (SELM), this paper proposes a Stacking Ensemble Learning Model based on One Class Classification (OCSELM) algorithms. In order to verify the effectiveness of the proposed algorithms, KDDCUP99 dataset, NSL-KDD dataset, UNSW-NB15 dataset and ISCX-URL2016 dataset are used for experiments. The experimental results show the superiority of the proposed algorithms.