Biblio

Website fingerprinting attack is a kind of traffic analysis attack that aims to identify the URL of visited websites using the Tor browser. Previous website fingerprinting attacks were based on batch learning methods which assumed that the traffic traces of each website are independent and generated from the stationary probability distribution. But, in realistic scenarios, the websites' concepts can change over time (dynamic websites) that is known as concept drift. To deal with data whose distribution change over time, the classifier model must update its model permanently and be adaptive to concept drift. Streaming algorithms are dynamic models that have these features and lead us to make a comparison of various representative data stream classification algorithms for website fingerprinting. Given to our experiments and results, by considering streaming algorithms along with statistical flow-based network traffic features, the accuracy grows significantly.

Malware evolves perpetually and relies on increasingly so- phisticated attacks to supersede defense strategies. Data-driven approaches to malware detection run the risk of becoming rapidly antiquated. Keeping pace with malware requires models that are periodically enriched with fresh knowledge, commonly known as retraining. In this work, we propose the use of Venn-Abers predictors for assessing the quality of binary classification tasks as a first step towards identifying antiquated models. One of the key benefits behind the use of Venn-Abers predictors is that they are automatically well calibrated and offer probabilistic guidance on the identification of nonstationary populations of malware. Our framework is agnostic to the underlying classification algorithm and can then be used for building better retraining strategies in the presence of concept drift. Results obtained over a timeline-based evaluation with about 90K samples show that our framework can identify when models tend to become obsolete.