Review Article
A Critical Review of Artificial Intelligence Based Approaches in Intrusion Detection: A Comprehensive Analysis
Table 2
Critical review of deep learning based approaches in ID.
| Ref | Authors | Year | Cited by | DL approach | Accuracy (%) |
| [48] | Yin et al. | 2017 | 1323 | RNN-IDS | 97.09 | [49] | Vani | 2017 | 19 | LSTM based ensemble method | 92.3 | [50] | Wang et al. | 2017 | 428 | Hierarchical spatial-temporal features-based IDS (HAST-IDS) with CNN | 99.89 | [51] | Loukas et al. | 2017 | 211 | RNN | 86.9 | [52] | Shone et al. | 2018 | 1046 | NDAE | 97.85 | [53] | Lee et al. | 2018 | 51 | Autoencoder | 98.9 | [54] | Al-Qatf et al. | 2018 | 334 | Self-taught learning (STL)-IDS | 99.41 | [55] | Ding and Zhai | 2018 | 91 | CNN | 80.13 | [56] | Parampottupadam and Moldovann | 2018 | 27 | Deep learning H2O (binomial and multinomial models) | 99.98 | [57] | Xin et al. | 2018 | 756 | CNN | 99.41 | [58] | Faker and Dogdu | 2019 | 140 | DNN | 99.16 | [59] | Laqtib et al. | 2019 | 15 | CNN | 77 | [60] | Ge et al. | 2019 | 121 | FFNN | 82 | [61] | Khan et al. | 2019 | 237 | Two-stage deep learning (TSDL) model | 99.31 | [62] | Gurung et al. | 2019 | 80 | Auto-encoders | 87.2 | [63] | Su et al. | 2020 | 153 | BAT model | 84.25 | [64] | Gamage and Samarabandu | 2020 | 156 | ANN | 98.25 | [65] | Boukhalfa et al. | 2020 | 30 | LSTM | 99.93 | [66] | Shende and Thorat | 2020 | 8 | LSTM | 96.92 | [67] | Kocher and Kumar | 2021 | 28 | ANN | 99.4 | [68] | Mighan and Kahani | 2021 | 65 | ANN | 98.51 | [69] | Ashiku and Dagli | 2021 | 33 | DNN | 95.6 | [70] | Salih et al. | 2021 | 20 | Bayesian CNN | 99.3271 | [71] | Imrana et al. | 2021 | 68 | Bidirectional (BiDLSTM) | 94.26 | [72] | Otoum et al. | 2022 | 115 | DL-IDS | 99 | [73] | Nasir et al. | 2022 | 13 | DF-IDS | 99.9 | [74] | Jasim | 2022 | 23 | Deep belief networks (DBNs) | 99 | [75] | Akshay Kumaar et al. | 2022 | 5 | DL-based hybrid framework “ImmuneNet” | 99.2 | [76] | Houda et al. | 2022 | 13 | Explainable artificial intelligence (XAI) based DL framework | 99 | [77] | Chaganti et al. | 2023 | 0 | LSTM | 97.1 | [78] | Figueiredo et al. | 2023 | 0 | LSTM | 99 | [79] | Rizvi et al. | 2023 | 2 | 1D-dilated causal neural network (1D-DCNN) | 99.98 |
|
|