Can We Trust Your Explanations? Sanity Checks for Interpreters in Android Malware Analysis
نویسندگان
چکیده
With the rapid growth of Android malware, many machine learning-based malware analysis approaches are proposed to mitigate severe phenomenon. However, such classifiers opaque, non-intuitive, and difficult for analysts understand inner decision reason. For this reason, a variety explanation interpret predictions by providing important features. Unfortunately, results obtained in domain cannot achieve consensus general, which makes confused about whether they can trust results. In work, we propose principled guidelines assess quality five designing three critical quantitative metrics measure their stability, robustness, effectiveness. Furthermore, collect widely-used datasets apply on them two tasks, including detection familial identification. Based generated results, conduct sanity check terms metrics. The demonstrate that our help us obtain knowledge most typical malicious behaviors analysis.
منابع مشابه
Android Security Permissions - Can We Trust Them?
The popularity of the Android System in combination with the lax market approval process may attract the injection of malicious applications (apps) into the market. Android features a permission system allowing a user to review the permissions an app requests and grant or deny access to resources prior to installation. This system conveys a level of trust due to the fact that an app only has ac...
متن کاملSanity Checks in Formal Verification
One of the advantages of temporal-logic model-checking tools is their ability to accompany a negative answer to the correctness query by a counterexample to the satisfaction of the specification in the system. On the other hand, when the answer to the correctness query is positive, most model-checking tools provide no additional information. In the last few years there has been growing awarenes...
متن کاملChallenges in Android Malware Analysis
HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau...
متن کاملCan We Trust Trust?
In this concluding essay I shall try to reconstruct what seem to me the central questions about trust that the individual contributions presented in this volume raise and partly answer. In the first section, I briefly qualify the claim that there is a degree of rational cooperation that should but does not exist, and I shall give a preliminary indication of the importance of the beliefs we hold...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE Transactions on Information Forensics and Security
سال: 2021
ISSN: ['1556-6013', '1556-6021']
DOI: https://doi.org/10.1109/tifs.2020.3021924