Deep Temporal Synergy: A TCN - LSTM Model for Advanced Malware Detection

In recent years, with an increase of usage of digital services and the proliferation of malware software there is a risk of malware attacks corrupting data, stealing information, or committing other cybercrimes. The accuracy of malware detection is limited due to traditional models' inability to completely capture long- term dependencies in API call sequences (e.g., CNN, LSTM). This paper provides a quick glimpse of hybrid models that unify the efficiency of Temporal Convolutional Networks and Long Short-Term Memory in the detection of malware. The model is evaluated using a public dataset with some improvement in performance measures. Our findings suggest that the TCN-LSTM hybrid approach helps detect malware easily and reduces false positive rates. This model achieves a validation accuracy of 97%.