Mohammed Ayalew Belay

Submitted: 22 February 2026 You are already at the latest version Anomaly detection is crucial for maintaining the safety, reliability, and optimal performance of complex systems across diverse domains such as industrial manufacturing, cybersecurity, and autonomous systems. Conventional methods typically handle single data modalities, limiting their effectiveness in the multimodal and dynamic real-world environments.

5. Numerical Results and Discussion In our numerical experiments, we conducted extensive tests on 13 different MTSAD algorithms covering all three of the main MTSAD categories discussed in Section 2. The algorithms are PCA, IF [42], OC-SVM [40] (conventional methods); VAE [64], MLP-AE [62], CNN-AE [89], GRU [49], LSTM [48], LSTM-AE [90], MAD-GAN [68] (DNN-based methods); ConvLSTM [83], USAD [85] and DAGMM [84] (composite models).