RADE: resource-efficient supervised anomaly detection using decision tree-based ensemble methods

The capability to perform anomaly detection in a resource-constrained setting, such as an edge device or loaded server, is of increasing need due emerging on-premises computation constraints well security, privacy and profitability reasons. Yet, the size datasets often results current methods being too resource consuming, particular decision-tree based ensemble classifiers. To address this need, we present RADE—a new resource-efficient framework that augments standard classifiers constrained setting. key idea behind RADE first train small model sufficient correctly classify majority queries. Then, using only subsets training data, expert models for these fewer harder cases where at high risk making classification mistake. We implement scikit-learn classifier. Our evaluation indicates offers competitive capabilities compared while significantly improving memory footprint by up $$12\times $$ , training-time $$20\times time $$16\times .