Enhanced Malware Detection System Based on AI Integrated Deep Feature Engineering using LSTM gated hyper capsule CNN

Increasing technical development of software technologies are more incredible to work with fascinated programs. However malwares that affect the programs to stumps working process to create various problems. Because of different types of malwares creates cyber threads leads attacks in multiple ways without user knowledge to create problems. Most of the Traditional methods implement malware detection programs without concentrating the nature of feature dependencies cause lower detection accuracy due to inappropriate feature dimension produce higher false positives. To resolve this problem, propose an advanced deep feature engineering-based malware detection system LSTM gated hyper capsule convolutional neural network. Initially the malware detection dataset is taken to process with c-score normalization to standardize the actual margins presence in feature values. Then Malware Behavior Impact Rate (MBIR) is estimated by using decisive tree to identify affects feature margins. Then Linear Progressive Ant Lion Optimization (LPALO) is applied to select the important features. Finally the LSTM gated hyper capsule convolutional neural network is applied to find the malware type. Proposed system improve the detection accuracy in higher true positive rate with improved precision, recall, f1 score with redundant false positive measures compared with existing methods. Experimentation on benchmark datasets demonstrates a remarkable improvement in detection rates over traditional approaches, highlighting the potential of deep feature learning in complementing current cyber security frameworks. The outcome demonstrates proposed system achieves high malware detection accuracy than traditional approaches.