• Publications
  • Influence
Tight Bounds on Minimax Regret under Logarithmic Loss via Self-Concordance
It is shown that for any expert class with (sequential) metric entropy, the minimax regret is $\mathcal{O}(n^{p/(p+1)})$, and that this rate cannot be improved without additional assumptions on the expert class under consideration.
Simulated co-location of patients admitted to an inpatient internal medicine teaching unit: potential impacts on efficiency and physician-nurse collaboration
A simulation model of inpatient flow through the internal medicine unit is developed and the impact of two proposed changes are determined: co-locating each team’s patients, and new admission rules for how patients are assigned to those teams.
Relaxing the I.I.D. Assumption: Adaptive Minimax Optimal Sequential Prediction with Expert Advice
It is shown that Hedge is suboptimal between these extremes, and a new algorithm is presented that is adaptively minimax optimal with respect to the authors' relaxations of the I.I.D. setting.
Relaxing the I.I.D. Assumption: Adaptively Minimax Optimal Regret via Root-Entropic Regularization
We consider sequential prediction with expert advice when data are generated from distributions varying arbitrarily within an unknown constraint set. We quantify relaxations of the classical i.i.d.
Stochastic Convergence Rates and Applications of Adaptive Quadrature in Bayesian Inference
Abstract We provide the first stochastic convergence rates for a family of adaptive quadrature rules used to normalize the posterior distribution in Bayesian models. Our results apply to the uniform
Average Waiting Times in the Two-Class M/G/1 Delayed Accumulating Priority Queue
Previously, Mojalal et al. (2019) gave an expression for the waiting time distribution of low priority customers in the Delayed Accumulating Priority Queue, but with no quantification of the effect
Minimax Optimal Quantile and Semi-Adversarial Regret via Root-Logarithmic Regularizers
This work extends existing KL regret upper bounds to possibly uncountable expert classes with arbitrary priors; provides the first full-information lower bounds for quantile regret on finite expert classes (which are tight); and provides an adaptively minimax optimal algorithm for the semi-adversarial paradigm that adapts to the true, unknown constraint faster, leading to uniformly improved regret bounds over existing methods.
Minimax Rates for Conditional Density Estimation via Empirical Entropy
We consider the task of estimating a conditional density using i.i.d. samples from a joint distribution, which is a fundamental problem with applications in both classification and uncertainty
High-Priority Expected Waiting Times in the Delayed Accumulating Priority Queue with Applications to Health Care KPIs
We provide the first analytical expressions for the expected waiting time of highpriority customers in the delayed APQ by exploiting a classical conservation law for work-conserving queues.