چکیده مقاله
In the domain of cybersecurity, the prompt and accurate detection of web attacks is of critical importance This paper presents a comparative analysis of two bidirectional Long Short Term Memory BiLSTM models, BiLSTM_v1 and BiLSTM_v2, for detecting web based attacks using the CSIC 2010 dataset Both models employ advanced preprocessing techniques and synthetic data balancing via the SMOTE algorithm to enhance their performance The BiLSTM_v1 model incorporates spatial dropout and batch normalization layers, whereas the BiLSTM_v2 model integrates additional dropout layers to potentially mitigate overfitting The training and validation phases were rigorously executed, with performance metrics such as accuracy, precision, recall, F1 score, and AUC Area Under Curve serving as the primary evaluation criteria The BiLSTM_v1 model achieved a test accuracy of 98 31% and an AUC of 99 91%, while the BiLSTM_v2 model demonstrated a slightly superior performance with a test accuracy of 98 43% and an AUC of 99 92% These results suggest that the architecture of BiLSTM_v2 offers marginally better performance in web attack detection The findings underscore the efficacy of BiLSTM networks in cybersecurity applications and provide a foundation for further advancements in intrusion detection systems
کلیدواژهها
نویسندگان
Salim Yarian۱
Master of Cybersecurity Imam Hossein Comprehensive University, Tehran
شیوه ارجاع
Yarian۱, Salim,1403,Evaluating BiLSTM_v1 and BiLSTM_v2: A Comparative Study of Bidirectional Long Short-Term Memory Networks for Web Attack Detection,The 22nd National Conference of Electrical, Computer and Mechanical Engineering,Shirvan
ارائهشده در
مجموعه مقالات بیست و دومین کنفرانس ملی مهندسی برق،کامپیوتر و مکانیک31 تیر 1403 · شیروان