Deep Learning (DL) Dense Classifier with Long Short-Term Memory Encoder Detection and Classification against Network Attacks

Abstract

The increase in network attacks is a serious threat. For the purpose of supporting study in connected areas, researchers provide datasets for network assaults. Following that, other attack detection techniques based on these datasets are suggested. In this paper, we have proposed a Deep Learning (DL) model called DL Dense Classifier with Long Short-Term Memory (LSTM) which is combination of Encoder and classifier modules to identify the type of network attacks from a massive malware dataset with 12 key malware categories and 141 features. Moreover, we have evaluated 3 classifier models on the CCCS-CIC-AndMal-2020 dataset utilizingprecision, recall, F-score, and accuracy metrics. The results of the experiments show that adding an LSTM encoder module to the DL Dense Classifier model successfully improves the accuracy of DL Dense Classifier from %87 to %91 and perform same as the ensemble model of Random Forest, which is one of the best ranked algorithms in the other research.

More by Selcuk Cankurt

• Effects of Binary Similarity Metrics on Energy Consumption During Recommendation by Using Collaborative Filtering ApproachJournal of sustainable development issues. · 2026
• STUDENT PERFORMANCE PREDICTION USING MACHINE LEARNINGZenodo · 2026
• Interpretable Hybrid CNN-BiLSTM-Attention Model for ECG Arrhythmia ClassificationIEEE International Conference on Smart Information Systems and Technologies · 2026
• COLLABORATIVE MULTI-AGENT CHATBOT FRAMEWORK FOR INTELLIGENT TUTORING AND FEEDBACKZenodo · 2026