Guest
Prof. Alexey Zaikin — Chair of Systems Medicine at UCL, physicist by training, h‑index 41 — explains how to extract reliable predictions from ultra‑high‑dimensional yet small clinical datasets. The lecture introduces Synolitic AI: * graphs grown solely from “healthy / diseased” labels, * ensembles of low‑dimensional classifiers that outperform giant LLMs, * noise‑resistant Synolitic Graph Neural Networks suitable for explainable medicine. Real‑world cases include COVID‑19 survival forecasting, early cancer detection, and age‑related trajectories in Down syndrome. Prof. Zaikin shows why Synolitic networks already beat classical methods and how they can merge with GAN‑based imaging and future graph‑augmented LLMs.
