Hybrid Intrusion Detection using MapReduce based Black Widow Optimized Convolutional Long Short-Term Memory Neural Networks
作者：
Highlights：
• IDS using Deep Learning algorithms improve Feature learning and reduce complexity.
• Built IDS using MapReduce based Black Widow Optimized Conv-LSTM Neural Network.
• Artificial Bee Colony based Feature selection reduced dimensionality.
• NSL-KDD, ISCX-IDS, UNSWNB15 and CSE-CIC-IDS2018 datasets are used for evaluation.
• Hyper parameter tuning in BWO-CONVLSTM increased accuracy and reduced complexity.
摘要
•IDS using Deep Learning algorithms improve Feature learning and reduce complexity.•Built IDS using MapReduce based Black Widow Optimized Conv-LSTM Neural Network.•Artificial Bee Colony based Feature selection reduced dimensionality.•NSL-KDD, ISCX-IDS, UNSWNB15 and CSE-CIC-IDS2018 datasets are used for evaluation.•Hyper parameter tuning in BWO-CONVLSTM increased accuracy and reduced complexity.
论文关键词：Intrusion Detection Systems,Deep Learning,Black Widow Optimization,Convolutional-Long Short-Term memory,Hyper-Parameter Optimization,NSL-KDD,ISCX-IDS,UNSW-NB15,CSE-CIC-IDS2018
论文评审过程：Received 14 May 2021, Revised 19 October 2021, Accepted 11 January 2022, Available online 22 January 2022, Version of Record 26 January 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.116545
