Tag
QDSP is an interpretable structured learning framework for predicting death or cerebral palsy in very low birth weight infants, integrating Quota-guided Subspace Sampling and Differentiable-decision-guided Structure Perception. It outperforms baselines like XGBoost, TabNet, and TabPFN on a real cohort and external datasets, identifying clinically relevant predictors such as cPVL and birth weight.
PRAXIS is a new algorithm that efficiently approximates the Rashomon set of near-optimal decision trees, achieving orders of magnitude improvement in runtime and memory while maintaining near-perfect recall.
This paper proposes a cost-effective UAV detection and classification system using sound signals processed by rational Gaussian wavelet neural networks, achieving interpretable and robust performance for single and multiple UAVs including swarms, outperforming traditional methods.