English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The popularity of Android OS has made it an appealing target to malwaredevelopers. To evade detection, including by ML-based techniques, attackersinvest in creating malware that closely resemble legitimate apps. In thispaper, we propose GUIDED RETRAINING, a supervised representation learning-basedmethod that boosts the performance of a malware detector. First, the dataset issplit into "easy" and "difficult" samples, where difficulty is associated tothe prediction probabilities yielded by a malware detector: for difficultsamples, the probabilities are such that the classifier is not confident on thepredictions, which have high error rates. Then, we apply our GUIDED RETRAININGmethod on the difficult samples to improve their classification. For the subsetof "easy" samples, the base malware detector is used to make the finalpredictions since the error rate on that subset is low by construction. For thesubset of "difficult" samples, we rely on GUIDED RETRAINING, which leveragesthe correct predictions and the errors made by the base malware detector toguide the retraining process. GUIDED RETRAINING focuses on the difficultsamples: it learns new embeddings of these samples using Supervised ContrastiveLearning and trains an auxiliary classifier for the final predictions. Wevalidate our method on four state-of-the-art Android malware detectionapproaches using over 265k malware and benign apps, and we demonstrate thatGUIDED RETRAINING can reduce up to 40.41% prediction errors made by the malwaredetectors. Our method is generic and designed to enhance the classificationperformance on a binary classification task. Consequently, it can be applied toother classification problems beyond Android malware detection.

A two-steps approach to improve the performance of Android malware detectors