Transitions, Losses, and Re-parameterizations: Elements of Prediction Games
This work offers foundational insights for researchers in machine learning, though it appears incremental as it builds on existing prediction game frameworks.
The thesis provides geometric insights into three types of two-player prediction games, addressing problems like general learning tasks and online convex optimization to understand intrinsic barriers and design efficient algorithms with theoretical guarantees such as constant regret.
This thesis presents some geometric insights into three different types of two player prediction games -- namely general learning task, prediction with expert advice, and online convex optimization. These games differ in the nature of the opponent (stochastic, adversarial, or intermediate), the order of the players' move, and the utility function. The insights shed some light on the understanding of the intrinsic barriers of the prediction problems and the design of computationally efficient learning algorithms with strong theoretical guarantees (such as generalizability, statistical consistency, and constant regret etc.).