Abstract:We study the problem of making predictions of an adversarially chosen high-dimensional state that are unbiased subject to an arbitrary collection of conditioning events, with the goal of tailoring these events to downstream decision makers. We give efficient algorithms for solving this problem, as well as a number of applications that stem from choosing an appropriate set of conditioning events. For example, we can efficiently make predictions targeted at polynomially many decision makers, giving each of them optimal swap regret if they best-respond to our predictions. We generalize this to online combinatorial optimization, where the decision makers have a very large action space, to give the first algorithms offering polynomially many decision makers no regret on polynomially many subsequences that may depend on their actions and the context. We apply these results to get efficient no-subsequence-regret algorithms in extensive-form games (EFGs), yielding a new family of regret guarantees for EFGs that generalizes some existing EFG regret notions, e.g. regret to informed causal deviations, and is generally incomparable to other known such notions. Next, we develop a novel transparent alternative to conformal prediction for building valid online adversarial multiclass prediction sets. We produce class scores that downstream algorithms can use for producing valid-coverage prediction sets, as if these scores were the true conditional class probabilities. We show this implies strong conditional validity guarantees including set-size-conditional and multigroup-fair coverage for polynomially many downstream prediction sets. Moreover, our class scores can be guaranteed to have improved $L_2$ loss, cross-entropy loss, and generally any Bregman loss, compared to any collection of benchmark models, yielding a high-dimensional real-valued version of omniprediction.

A Universal Probability Assignment for Prediction of Individual Sequences

Sequential prediction of individual sequences under general loss functions

Predictability and Randomness

Predictions and algorithmic statistics for infinite sequence

Characterizing predictable classes of processes

High-Dimensional Prediction for Sequential Decision Making

A Family of LZ78-based Universal Sequential Probability Assignments

Tracking Changing Probabilities via Dynamic Learners

On Universal Prediction and Bayesian Confirmation

+ YY ’ Z X-Residual Prediction Error TargetPredictionObservation Latent 0 State φ f 1 f 2

Binomial Prediction Using the Frequent Outcome Approach

Random scattering of bits by prediction

Universal time-series forecasting with mixture predictors

Prediction with eventual almost sure guarantees

Prediction from compression for models with infinite memory, with applications to hidden Markov and renewal processes

Probability-Matching Predictors for Extreme Extremes

On Confidence Sequences for Bounded Random Processes via Universal Gambling Strategies

Studies in subjective probability l: Prediction of random events

Learning Predictions for Algorithms with Predictions

Universal Supervised Learning for Individual Data

Learning the Experts for Online Sequence Prediction