AI for Health #24 Making Neural Nets More Causally Accurate for Clinical Applications, David Page скачать в хорошем качестве

AI for Health #24 Making Neural Nets More Causally Accurate for Clinical Applications, David Page

In this AI for Health Webinar, Dr. David Page (James B. Duke Distinguished Professor & Chair of Biostatistics & Bioinformatics, Duke University) presents his research on making neural networks more causally accurate for clinical applications. He discusses why causal accuracy matters in healthcare AI, approaches such as matching, propensity-score methods, and causal transformers, and new theoretical contributions including a model of probably approximately correct causal discovery and a formal connection between neural networks and probabilistic graphical models. Dr. Page’s work focuses on algorithms for data mining and machine learning applied to biomedical data, especially de-identified electronic health records, high-throughput genetic and molecular data, and complex multi-relational data. He explores methods for finding causal relationships and producing human-interpretable outputs that can improve healthcare AI. 📌 Recorded as part of the AI for Health Webinar Series Speaker lineup: https://sites.google.com/view/ai4health/ Past recordings:    • AI for health webinar   #AIforHealth #CausalAI #HealthcareAI #NeuralNetworks #ClinicalAI #KeijiAI