The purpose of this study was to explore cultural similarities and differences in regret, focusing on distinctions between interpersonal and self-situations, and between action and inaction regrets. Japanese and American undergraduates were asked to describe regrets experienced in interpersonal and self-situations. We found that both situational and cultural contexts influenced the likelihood o...

We consider regret minimization in adversarial deterministic Markov Decision Processes (ADMDPs) with bandit feedback. We devise a new algorithm that pushes the state-of-theart forward in two ways: First, it attains a regret of O(T ) with respect to the best fixed policy in hindsight, whereas the previous best regret bound was O(T ). Second, the algorithm and its analysis are compatible with any...

Various papers have analyzed the noisy optimization of convex functions. This analysis has been made according to several criteria used to evaluate the performance of algorithms: uniform rate, simple regret and cumulative regret. We propose an iterative optimization framework, a particular instance of which, using Hessian approximations, provably (i) reaches the same rate as Kiefer-Wolfowitz al...

Classical stochastic programming has already been used with large scale LP models for long-term analysis of energy-environment systems. We propose a Minimax Regret formulation suitable for large scale linear programming models. It has been experimentally verified that the minimax regret strategy depends only on the extremal scenarios and not on the intermediate ones, thus making the approach co...

We introduce two online gradient-based reinforcement learning algorithms with function approximation – one model based, and the other model free – for which we provide a regret analysis. Our regret analysis has the benefit that, unlike many other gradient based algorithm analyses for reinforcement learning with function approximation, it makes no probabilistic assumptions meaning that we need n...

We analyze competition among newsvendors when the only information competitors possess about the nature of future demand realizations is the support of demand distributions. In such a setting, traditional expectation-based optimization criteria may not be adequate. In our analysis, we focus on several alternative criteria used in the robust optimization literature, such as relative and absolute...

We introduce a new algorithm and a new analysis technique that is applicable to a variety of online optimization scenarios, including regret minimization for Lipschitz regret functions, universal portfolio management, online convex optimization and online utility maximization. In addition to being more efficient and deterministic, our algorithm applies to a more general setting (e.g. when the p...

Which domains in life produce the greatest potential for regret, and what features of those life domains explain why? Using archival and laboratory evidence, the authors show that greater perceived opportunity within life domains evokes more intense regret. This pattern is consistent with previous publications demonstrating greater regret stemming from high rather than low opportunity or choice...

Several theories have been proposed to account for variation in the intensity of life regrets. Variables hypothesized to affect the intensity of regret include: whether the regretted decision was an action or an inaction, the degree to which the decision was justified, and the life domain of the regret. No previous study has compared the effects of these key predictors in a single model in orde...

UCT, a state-of-the art algorithm for Monte Carlo tree sampling (MCTS), is based on UCB, a sampling policy for the Multi-armed Bandit Problem (MAB) that minimizes the accumulated regret. However, MCTS differs from MAB in that only the final choice, rather than all arm pulls, brings a reward, that is, the simple regret, as opposite to the cumulative regret, must be minimized. This ongoing work a...