N. Mikaeilvand

A stable hierarchical deep learning architecture for cross-layer performance modeling in cloud computing

A hybrid graph- and sequence-based deep learning framework for behavioral analysis of metamorphic malware

This paper investigates a unified behavioral analysis framework that combines temporal and structural representations of Windows API call sequences: a sequential view based on learned API n-gram representations and a Bidirectional LSTM (BiLSTM) encoder, a structural view derived from the API transition graph, and a mul...

DeepOpTED: An intelligent deep operator network for text emotion recognition

This is the first application of DeepONet to textual emotion recognition, introducing a fundamentally different perspective on function approximation in language understanding tasks and suggesting that DeepONet can serve as a robust alternative framework for capturing and modeling complex relationships in natural langu...