Convolutional Neural Networks as Classification Tools and Feature Extractors for Distinguishing Malware Programs
| Title | Convolutional Neural Networks as Classification Tools and Feature Extractors for Distinguishing Malware Programs |
| Publication Type | Conference Paper |
| Year of Publication | 2019 |
| Authors | Priyamvada Davuluru, Venkata Salini, Narayanan Narayanan, Barath, Balster, Eric J. |
| Conference Name | 2019 IEEE National Aerospace and Electronics Conference (NAECON) |
| Date Published | July 2019 |
| Publisher | IEEE |
| ISBN Number | 978-1-7281-1416-3 |
| Keywords | anti-malware industry, classification tools, CNN based algorithms, computationally efficient CNN-based architecture, Computer architecture, convolutional neural nets, convolutional neural networks, feature extraction, feature extractors, Human Behavior, invasive software, k-nearest neighbors, KNN, learning (artificial intelligence), Malware, malware classification, malware detection, malware programs, Metrics, microsoft malware classification challenge, nearest neighbour methods, pattern classification, privacy, pubcrawl, resilience, Resiliency, support vector machine, Support vector machines, SVM, Testing, Training, visualization |
| Abstract | Classifying malware programs is a research area attracting great interest for Anti-Malware industry. In this research, we propose a system that visualizes malware programs as images and distinguishes those using Convolutional Neural Networks (CNNs). We study the performance of several well-established CNN based algorithms such as AlexNet, ResNet and VGG16 using transfer learning approaches. We also propose a computationally efficient CNN-based architecture for classification of malware programs. In addition, we study the performance of these CNNs as feature extractors by using Support Vector Machine (SVM) and K-nearest Neighbors (kNN) for classification purposes. We also propose fusion methods to boost the performance further. We make use of the publicly available database provided by Microsoft Malware Classification Challenge (BIG 2015) for this study. Our overall performance is 99.4% for a set of 2174 test samples comprising 9 different classes thereby setting a new benchmark. |
| URL | https://ieeexplore.ieee.org/document/9058025/ |
| DOI | 10.1109/NAECON46414.2019.9058025 |
| Citation Key | priyamvada_davuluru_convolutional_2019 |
- malware detection
- visualization
- Training
- testing
- SVM
- Support vector machines
- support vector machine
- Resiliency
- resilience
- pubcrawl
- privacy
- pattern classification
- nearest neighbour methods
- microsoft malware classification challenge
- Metrics
- malware programs
- anti-malware industry
- malware classification
- malware
- learning (artificial intelligence)
- KNN
- k-nearest neighbors
- invasive software
- Human behavior
- feature extractors
- feature extraction
- convolutional neural networks
- convolutional neural nets
- computer architecture
- computationally efficient CNN-based architecture
- CNN based algorithms
- classification tools