Part of the AI People directory: a structured reference to the individuals shaping artificial intelligence across research, industry, governance, ethics and public discourse.
Vladimir Vapnik, statistical learning theory, United States. Developed statistical learning theory and co-invented Support Vector Machines with Corinna Cortes. His VC theory provides the mathematical foundation for understanding when and why machine learning algorithms generalise from training data. *Key works: The Nature of Statistical Learning Theory (1995).*
Corinna Cortes, machine learning, United States. Co-invented Support Vector Machines with Vapnik, one of the most influential algorithms of the 1990s and 2000s, and led machine learning research at Google, applying these methods at unprecedented scale. *Key works: Support-Vector Networks (1995).*
Michael I. Jordan, probabilistic models and ML foundations, United States. One of the most cited researchers in machine learning, known for graphical models, variational inference and the mathematical foundations of ML. He trained many leading researchers, and his 2019 essay warning about AI hype remains widely referenced. *Key works: foundational papers on graphical models, variational methods and Bayesian approaches.*
Bernhard Schölkopf, kernel methods and causality, Germany. Pioneered kernel methods in machine learning and later became a leading voice on causality, arguing that understanding cause and effect, not just correlation, is essential for robust AI. Leads the Max Planck Institute for Intelligent Systems. *Key works: Learning with Kernels (2002); causality research.*
Leo Breiman, ensemble methods, United States. Created random forests and bagging, techniques that combine multiple models to improve accuracy. His 2001 paper Statistical Modeling: The Two Cultures articulated a tension between traditional statistics and machine learning that persists today. *Key works: Random Forests (2001); Statistical Modeling: The Two Cultures (2001).*
Daphne Koller, probabilistic models and AI in biology, United States. Co-authored the definitive textbook on probabilistic graphical models and co-founded Coursera, democratising AI education globally. Now leads Insitro, applying machine learning to drug discovery. *Key works: Probabilistic Graphical Models (2009); the Coursera platform.*
Christopher Bishop, Bayesian methods and ML education, United Kingdom. Wrote Pattern Recognition and Machine Learning, one of the most influential ML textbooks, and led Microsoft Research Cambridge, shaping how a generation learned the mathematical foundations of the field. *Key works: Pattern Recognition and Machine Learning (2006).*
Leslie Valiant, computational learning theory, United States. Created PAC (Probably Approximately Correct) learning theory, the first rigorous mathematical framework for understanding machine learning. A Turing Award winner whose theoretical work underpins our understanding of what is learnable. *Key works: A Theory of the Learnable (1984).*
Trevor Hastie, Robert Tibshirani and Jerome Friedman, statistical learning, United States. Co-authored The Elements of Statistical Learning, the canonical textbook bridging statistics and machine learning. Tibshirani invented LASSO regularisation; Friedman developed gradient boosting. *Key works: The Elements of Statistical Learning (2001); LASSO; gradient boosting.*
Yoav Freund and Robert Schapire, boosting, United States. Created AdaBoost, demonstrating that combining weak learners could produce strong predictive models. Their theoretical and practical work on boosting influenced machine learning competitions and production systems for two decades. *Key works: the AdaBoost algorithm and boosting theory.*
Kevin Murphy, probabilistic ML and education, United States. Author of Probabilistic Machine Learning, a comprehensive modern textbook that has become essential reading, helping thousands of practitioners understand the mathematical foundations of modern AI. *Key works: Probabilistic Machine Learning (2022).*