|
AUTHOR(S): Patricia Bossah, Bartholomew Idoko
|
TITLE Development of a Hybrid Machine Learning Model (Inception V3 +SVM) for Malware Detection |
 PDF |
ABSTRACT The malware landscape is constantly evolving. The tools used for malware detection and threat elimination in critical environments must evolve too. Discovering threats before they can do harm presents a continuous challenge to the world of cybersecurity. By gaining a better understanding of malware and how it functions, we can take proper steps to eliminate these threats from the systems we want to protect. The research considers; the practice of Malware analysis with threat hunting, Indicator of Attack (IoA) and Indicator of Compromise (IoC) and the Identification and prevention of threats when they would otherwise bypass the traditional detection systems using Machine Learning (ML) model. In this paper, a hybrid model is proposed as a framework that can integrate different ML techniques so as to better improve the detection and classification accuracy. The system is implemented using deep learning structure (Inception v3) plus SVM classifier. The framework’s input comes from instances of the explored malware dataset where these instances are fed into the hybrid model via the Inception v3 input layer. The instances are further forwarded to convolutional phase of Inception v3, where important features are extracted. The extracted features are utilized in form of support vector machine classifier’s input. This incorporated support vector machine classifier uses the extracted features for classification. The classification report of inception v3 + SVM model yields a better accuracy when compared to that of the SVM model. The advantage of this framework is that it can be applied to train and partition a very large data set over and over. This framework will guarantee a more accurate and robust detection system that could outperformed the conventional malware detection models. |
KEYWORDS Malware, detection; Machine Learning; accuracy; classification; support vector machine, inception v3 |
|
Cite this paper Patricia Bossah, Bartholomew Idoko. (2025) Development of a Hybrid Machine Learning Model (Inception V3 +SVM) for Malware Detection. International Journal of Signal Processing, 10, 23-33 |
|
