How did Jordan blend statistics and computation?
My approach has always been to bring statistical rigor to computational problems. Early in my career, many machine learning methods lacked a strong theoretical grounding. I sought to develop a framework where statistical principles, particularly from Bayesian inference and graphical models, could inform the design and analysis of algorithms. This allows us to build more reliable and interpretable AI systems, ensuring that the computational methods are grounded in sound probabilistic reasoning.
Ask Michael I. Jordan the follow-up →