arXiv research

Presents SPEED, an algorithm for optimal policy evaluation in linear bandits with heteroscedastic noise.

The paper analyzes the sample complexities for policy evaluation with linear function approximation.

SVRG reduces gradient evaluations for policy evaluation in reinforcement learning.

This work characterizes conditions for offline policy evaluation in reinforcement learning.

BCRL learns a Bellman complete representation for offline RL policy evaluation.

Unified coverage analysis for linear off-policy evaluation in reinforcement learning.

Bayesian optimization (BO) is a model-based approach for gradient-free black-box function optimization. Typically, BO is powered by a Gaussian process (GP), whose algorithmic complexity is cubic in the number of evaluations. Hence, GP-based BO cannot leverage large amounts of past or related function evaluations, for e…

The paper tackles noisy multi-armed bandit problems with improved regret guarantees.

VA-OPE improves OPE by incorporating variance information, achieving tighter error bounds.

The paper analyzes the generalizability of linear autoencoders and multivariate linear regression.

Improving the interpretability of brain decoding approaches is of primary interest in many neuroimaging studies. Despite extensive studies of this type, at present, there is no formal definition for interpretability of brain decoding models. As a consequence, there is no quantitative measure for evaluating the interpre…

The paper constructs semistrict monoidal 2-categories from foam evaluations.

Policy evaluation is a crucial step in many reinforcement-learning procedures, which estimates a value function that predicts states' long-term value under a given policy. In this paper, we focus on policy evaluation with linear function approximation over a fixed dataset. We first transform the empirical policy evalua…

We develop a theory for solving continuous time optimal stopping problems for non-linear expectations. Our motivation is to consider problems in which the stopper uses risk measures to evaluate future rewards.

New algorithm optimally evaluates policies with linear approximations.

LinMED is a new linear bandit algorithm with near-optimal regret bound.

Off-policy reinforcement learning has many applications including: learning from demonstration, learning multiple goal seeking policies in parallel, and representing predictive knowledge. Recently there has been an proliferation of new policy-evaluation algorithms that fill a longstanding algorithmic void in reinforcem…

Paper improves TD learning algorithm bounds with linear approx.

This research evaluates learning models for bionic robots, focusing on transfer function identification.

UCoS avoids forward model evaluations in sampling for large-scale linear inverse problems.

New method efficiently evaluates policies using trajectory data.

We present a method to stop the evaluation of a prediction process when the result of the full evaluation is obvious. This trait is highly desirable in prediction tasks where a predictor evaluates all its features for every example in large datasets. We observe that some examples are easier to classify than others, a p…

Paper develops a method to highlight influential transitions for off-policy evaluation.

In this paper we investigate the performance of different types of rectified activation functions in convolutional neural network: standard rectified linear unit (ReLU), leaky rectified linear unit (Leaky ReLU), parametric rectified linear unit (PReLU) and a new randomized leaky rectified linear units (RReLU). We evalu…

Adequate evaluation of an information retrieval system to estimate future performance is a crucial task. Area under the ROC curve (AUC) is widely used to evaluate the generalization of a retrieval system. However, the objective function optimized in many retrieval systems is the error rate and not the AUC value. This p…

The paper evaluates joint life insurance risk under dependence uncertainty using copulas and convex risk measures.

Improved electricity price forecasting model combining linear and non-linear structures.

The paper simplifies pricing for equity swaps by accounting for various costs.

Cross-validation (CV) is a technique for evaluating the ability of statistical models/learning systems based on a given data set. Despite its wide applicability, the rather heavy computational cost can prevent its use as the system size grows. To resolve this difficulty in the case of Bayesian linear regression, we dev…

Study shows PI for LQR requires fewer policy improvement steps than policy evaluation steps.

Improves comparison of F-measures for imbalanced datasets.

New methods combine low and high-fidelity data for accurate surrogate modeling.

A new estimator reduces variance in slate bandit OPE.

Fast algorithms developed for adaptive and fully adaptive submodular maximization problems.

New bounds for nearly-linear networks without training.

Proposes a modified Morgan-Pitman test for evaluating variances in machine learning models.

In order to obtain a reasonable and reliable forecast method for crude oil price volatility, this paper evaluates the forecast performance of single-regime GARCH models (including the standard linear GARCH model and the nonlinear GJR-GARCH and EGARCH models) and the two-regime Markov Regime Switching GARCH (MRS-GARCH) …

New dataset and analysis improve evaluation of visual representation models.

In this pre-print we explore the multi-fractal properties of 1 minute traded volume of the equities which compose the Dow Jones 30. We also evaluate the weights of linear and non-linear dependences in the multi-fractal structure of the observable. Our results show that the multi-fractal nature of traded volume comes es…

The study sets lower bounds on MMSE for inferring sensitive features from noisy data.

Paper introduces Modular Jets for diagnosing model decompositions in pipelines.

New algorithm reduces multi-agent bandit regret by sharing data.

Sparse grids reduce xVA exposure evaluations by up to 6000 times.

The paper develops predictors for functional data on manifolds.

KBB algorithm reduces sample complexity for policy evaluation in general state spaces.

Policy gradient methods achieve linear convergence in simple MDPs.

SELO model predicts link signs better than SDGNN using subgraph encoding and linear optimization.

Paper optimizes broker performance by estimating execution costs.