Imbalanced Malware Images Classification: a CNN based Approach

Abstract

Deep convolutional neural networks (CNNs) can be applied to malware binary detection via image classification. The performance, however, is degraded due to the imbalance of malware families (classes). To mitigate this issue, we propose a simple yet effective weighted softmax loss which can be employed as the final layer of deep CNNs. The original softmax loss is weighted, and the weight value can be determined according to class size. A scaling parameter is also included in computing the weight. Proper selection of this parameter is studied and an empirical option is suggested. The weighted loss aims at alleviating the impact of data imbalance in an end-to-end learning fashion. To validate the efficacy, we deploy the proposed weighted loss in a pre-trained deep CNN model and fine-tune it to achieve promising results on malware images classification. Extensive experiments also demonstrate that the new loss function can well fit other typical CNNs, yielding an improved classification performance.

updated version with typos fixed and corrected technical expressions

References in corpus (5)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Deep Neural Network Based Malware Detection Using Two Dimensional Binary Program Features
• Dilated Deep Residual Network for Image Denoising
• An ELU Network with Total Variation for Image Denoising
• Deep residual learning in CT physics: scatter correction for spectral CT

Cited by in corpus (13)

• Microsoft Malware Classification Challenge
• A Few-Shot Meta-Learning based Siamese Neural Network using Entropy Features for Ransomware Classification
• Deep Multi-Task Learning for Malware Image Classification
• Lightweight Classification of IoT Malware based on Image Recognition
• Generation & Evaluation of Adversarial Examples for Malware Obfuscation
• Deep Transfer Learning for Static Malware Classification
• A Visualized Malware Detection Framework with CNN and Conditional GAN
• Malware Detection Using Frequency Domain-Based Image Visualization and Deep Learning
• Towards Interpretable Ensemble Learning for Image-based Malware Detection
• Understanding the efficacy, reliability and resiliency of computer vision techniques for malware detection and future research directions
• Sequential Embedding-based Attentive (SEA) classifier for malware classification
• To believe or not to believe: Validating explanation fidelity for dynamic malware analysis
• A Review of Computer Vision Methods in Network Security