Machine Learning-Based Malware Software Detection Based on Adaptive Gradient Support Vector Regression

Machine Learning-Based Malware Software Detection Based on Adaptive Gradient Support Vector Regression Lavanya Bharathi Shanthi Chandrabose

Malware Software detection is one of the key steps in developing the anti-malware software in computer systems. In the existing system, malware detection had been performed inefficiently with poor detection accuracy. The previous methods were not efficient enough to detect malware in terms of low efficiency, low overhead, and poor security. The proposed method uses the Machine learning approaches for Malware software detection based on the Adaptive Gradient Support Vector Regression (AGSVR) to overcome these issues. Initially, the pre-processing stage reduces the imbalanced data and missing values based on the Adaptive Normalized Data Analysis (ANDA) using the specified dataset. Secondly, features extracted from the pre-processing stage are used for the training and testing of dataset using the Adaptive Static Feature Analysis (ASFA) algorithm. Each selected feature value is extracted and stored with the associated category of specified dataset. Absolute rights are established based on the values assigned to the Malware software detection system. Finally, the analysis of the selected features is done using the classification based on the training and testing of malware data. The classification is based on the Adaptive Gradient Support Vector Regression (AGSVR) algorithm. Recognition is an approach to mutual identification that is useful for distinguishing between malicious and non-malicious applications. Then, the extracted information is used to classify malicious and benign applications that use machine learning-based AGSVR classification algorithm. The simulation results show the improved sensitivity and specificity, reduced error rate, high accuracy and reduced time complexity in the proposed method which is better than the previous method.
02 28 2022 02 28 2022 39 45 Crossmark v2.0 10.18280/CrossmarkPolicy www.iieta.org true 30 November 2021 15 January 2022 28 February 2022 http://iieta.org/sites/default/files/TEXT%20AND%20DATA%20MINING%20SERVICE%20AGREEMENT.pdf 10.18280/ijsse.120105 https://www.iieta.org/journals/ijsse/paper/10.18280/ijsse.120105 https://www.iieta.org/journals/ijsse/paper/10.18280/ijsse.120105