arXiv research

Optimistic NPG improves policy optimization in online RL with efficient sample complexity.

Omega method mitigates noise in stochastic game optimization.

New algorithms reduce forecasting errors by leveraging optimistic learning and hinting.

Proposes Optimistic Pessimistically Initialised Q-Learning (OPIQ) for better exploration in RL.

New algorithm reduces regret in sequential decision-making problems.

Optimistic method adapted for faster convex-concave min-max problems.

Unified framework for model- and value-optimistic reinforcement learning.

Accelerates policy optimization in RL with optimistic and adaptive updates.

This paper develops a model of reference-dependent assessment of subjective beliefs in which loss-averse people optimally choose the expectation as the reference point to balance the current felicity from the optimistic anticipation and the future disappointment from the realisation. The choice of over-optimism or over…

Optimistic Hedge achieves optimal regret bounds in two-player zero-sum games.

A fundamental problem arising in many areas of machine learning is the evaluation of the likelihood of a given observation under different nominal distributions. Frequently, these nominal distributions are themselves estimated from data, which makes them susceptible to estimation errors. We thus propose to replace each…

New method for optimistic planning in MDPs using regularization.

Efficiently solves exploration-exploitation in LQR using Lagrangian relaxation.

Develops ODRPO to improve RL algorithms with better performance and stability.

Optimistic RL algorithms are simplified for deep RL with competitive performance.

Paper introduces DOO models to outperform SAA out-of-sample.

Optimistic PPO variant solves linear MDPs with improved regret bound.

Optimistic pricing algorithm handles online dynamic pricing with censored demand.

It is well known that in stochastic multi-armed bandits (MAB), the sample mean of an arm is typically not an unbiased estimator of its true mean. In this paper, we decouple three different sources of this selection bias: adaptive \emph{sampling} of arms, adaptive \emph{stopping} of the experiment, and adaptively \emph{…

New RL algorithm optimizes policies with bandit feedback, matching previous bounds.

The likelihood function is a fundamental component in Bayesian statistics. However, evaluating the likelihood of an observation is computationally intractable in many applications. In this paper, we propose a non-parametric approximation of the likelihood that identifies a probability measure which lies in the neighbor…

Robust Optimization has traditionally taken a pessimistic, or worst-case viewpoint of uncertainty which is motivated by a desire to find sets of optimal policies that maintain feasibility under a variety of operating conditions. In this paper, we explore an optimistic, or best-case view of uncertainty and show that it …

PROPO tackles non-stationary MDPs with efficient policy optimization.

Optimistic algorithm reduces regret and constraint violations in online convex optimization with adversarial constraints.

Proposes H-UCRL for efficient model-based RL with sublinear regret.

Study optimizes linear regression analysis for high-dimensional settings.

StoSOO optimistically maximizes noisy, locally smooth functions.

Model-based Bayesian Reinforcement Learning (BRL) allows a found formalization of the problem of acting optimally while facing an unknown environment, i.e., avoiding the exploration-exploitation dilemma. However, algorithms explicitly addressing BRL suffer from such a combinatorial explosion that a large body of work r…

Optimizes algorithms for non-concave bandit problems.

Optimistic bounds for multi-output learning using self-bounding Lipschitz condition.

Bayesian classifier improves robustness with optimistic score ratio.

New learning rules achieve optimal sample complexity for weakly supervised classification.

Improved regret bounds for online convex optimization under stochastic and adversarial settings.

We show that the optimistic limits of the colored Jones polynomials of the hyperbolic knots coincide with the optimistic limits of the Kashaev invariants modulo $4π^2$.

We address the issue of limit cycling behavior in training Generative Adversarial Networks and propose the use of Optimistic Mirror Decent (OMD) for training Wasserstein GANs. Recent theoretical results have shown that optimistic mirror decent (OMD) can enjoy faster regret rates in the context of zero-sum games. WGANs …

We discuss the relative merits of optimistic and randomized approaches to exploration in reinforcement learning. Optimistic approaches presented in the literature apply an optimistic boost to the value estimate at each state-action pair and select actions that are greedy with respect to the resulting optimistic value f…

New RL algorithm tackles online robust MDPs with uncertainty.

Optimistic search speeds up change point detection in large datasets.

The optimistic limit is the mathematical formulation of the classical limit which is a physical method to expect the actual limit by using saddle point method of certain potential function. The original optimistic limit of the Kashaev invariant was formulated by Yokota, and a modified formulation was suggested by the a…

New algorithm achieves data-dependent regret bounds in MDPs with unknown transitions.

We consider the problem of minimizing a smooth convex function by reducing the optimization to computing the Nash equilibrium of a particular zero-sum convex-concave game. Zero-sum games can be solved using online learning dynamics, where a classical technique involves simulating two no-regret algorithms that play agai…

Optimistic estimate predicts best fitting performance of nonlinear models.

We propose a new variant of AMSGrad, a popular adaptive gradient based optimization algorithm widely used for training deep neural networks. Our algorithm adds prior knowledge about the sequence of consecutive mini-batch gradients and leverages its underlying structure making the gradients sequentially predictable. By …

New algorithm AG-OG optimizes separable convex-concave problems efficiently.

New algorithm reduces reinforcement learning regret to sqrt(d^3T).

Characterizes concept classes for optimistic online learning.

OPPO optimizes policy learning with provable efficiency.

New algorithms reduce regret in online MDPs by adapting to data and variance.