===================================== Regret Bounds for Prediction Problems ===================================== - Geoff Gordon We present a unified framework for reasoning about worst-case regret bounds for learning algorithms. This framework is based on the theory of duality of convex functions. It brings together results from computational learning theory and Bayesian statistics, allowing us to derive new proofs of known theorems, new theorems about known algorithms, and new algorithms.