On the Limitations of Continual Learning for Malware Classification-猫眼课题宝

DOI：

10.48550/arxiv.2208.06568

发表时间：

2022-08

期刊：

ArXiv

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Mohammad Saidur Rahman;Scott E. Coull;M. Wright

恶意软件(Malware)分类为持续学习(CL)机制带来了独特的挑战，因为每天都会收到大量新样本，并且恶意软件不断演变以利用新的漏洞。在典型的一天里，反病毒供应商会收到数十万个独特的恶意和良性软件，在恶意软件分类器的整个生命周期中，可以轻松积累超过10亿个样本。考虑到问题的规模，使用持续学习技术的顺序培训可以在减少培训和存储开销方面提供实质性的好处。然而，到目前为止，还没有关于CL应用于恶意软件分类任务的探索。本文研究了11种CL技术应用于三种常见的增量学习场景的恶意软件任务，包括任务、类和域增量学习(IL)。具体地说，使用两个真实的大规模恶意软件数据集，我们评估了CL方法在二进制恶意软件分类(域-IL)和多类恶意软件家族分类(任务-IL和类-IL)任务上的性能。令我们惊讶的是，在几乎所有的设置中，持续学习方法在训练数据的朴素联合回放方面都表现得很差--在某些情况下，准确率降低了70多个百分点。选择性地重放20%的存储数据的简单方法获得了更好的性能，与联合重放相比，训练时间减少了50%。最后，我们讨论了CL技术性能出乎意料地差的潜在原因，希望这能促进进一步研究开发在恶意软件分类领域更有效的技术。

Malicious software (malware) classification offers a unique challenge for continual learning (CL) regimes due to the volume of new samples received on a daily basis and the evolution of malware to exploit new vulnerabilities. On a typical day, antivirus vendors receive hundreds of thousands of unique pieces of software, both malicious and benign, and over the course of the lifetime of a malware classifier, more than a billion samples can easily accumulate. Given the scale of the problem, sequential training using continual learning techniques could provide substantial benefits in reducing training and storage overhead. To date, however, there has been no exploration of CL applied to malware classification tasks. In this paper, we study 11 CL techniques applied to three malware tasks covering common incremental learning scenarios, including task, class, and domain incremental learning (IL). Specifically, using two realistic, large-scale malware datasets, we evaluate the performance of the CL methods on both binary malware classification (Domain-IL) and multi-class malware family classification (Task-IL and Class-IL) tasks. To our surprise, continual learning methods significantly underperformed naive Joint replay of the training data in nearly all settings -- in some cases reducing accuracy by more than 70 percentage points. A simple approach of selectively replaying 20% of the stored data achieves better performance, with 50% of the training time compared to Joint replay. Finally, we discuss potential reasons for the unexpectedly poor performance of the CL techniques, with the hope that it spurs further research on developing techniques that are more effective in the malware classification domain.

On the Limitations of Continual Learning for Malware Classification

文献摘要

相似文献