Speaker
Description
Machine Learning (ML) adoption within large-scale networks is limited. Organisations hesitate to share data in fear of compromising end-user privacy, thus representative datasets to train accurate ML classifiers are usually not available. Moreover, black-box classifiers are not intrinsically explainable, hence engineers are reluctant to deploy them. FedXAI4DNS employs ML, Federated Learning (FL) and eXplainable AI (XAI) for collaborative and trustworthy detection of malignant DNS traffic produced by Domain Generation Algorithms (DGAs). FL enables collaborators to train classifiers without exchanging sensitive data, whereas XAI analyses black-box model operation. FedXAI4DNS aims at expediting ML adoption within collaborative environments (e.g. NRENs & GÉANT).
