arXiv research

Unified algorithm for reinforcement learning with function approximation.

A significant hurdle for analyzing large sample data is the lack of effective statistical computing and inference methods. An emerging powerful approach for analyzing large sample data is subsampling, by which one takes a random subsample from the original full sample and uses it as a surrogate for subsequent computati…

SLIM efficiently solves overidentified models in a scalable manner.

The paper shows how sketching data can simplify regression inference even when errors are heteroskedastic.

Efficiently learns Ising model parameters with limited statistics.

Subbagging estimation for big data reduces memory usage while maintaining statistical consistency.

The paper analyzes Karcher means on restricted PSD matrices with statistical guarantees.

Develops bounds for deep learning risk via Hilbert coresets.

In certain situations that shall be undoubtedly more and more common in the Big Data era, the datasets available are so massive that computing statistics over the full sample is hardly feasible, if not unfeasible. A natural approach in this context consists in using survey schemes and substituting the "full data" stati…

This paper analyzes the direction of the causality between crude oil, gold and stock markets for the largest economy in the world with respect to such markets, the US. To do so, we apply non-linear Granger causality tests. We find a nonlinear causal relationship among the three markets considered, with the causality go…

A new method for kernel tests without data splitting increases power.

We propose a generic Markov Chain Monte Carlo (MCMC) algorithm to speed up computations for datasets with many observations. A key feature of our approach is the use of the highly efficient difference estimator from the survey sampling literature to estimate the log-likelihood accurately using only a small fraction of …

Meta-learners improve causal effect estimation in small samples.

New method filters large networks from financial data to reveal key subnetworks.

The paper analyzes error bounds and KL properties for noisy matrix recovery problems.

This paper shows using sub-sample estimates can improve optimization results in large-scale problems.

Efficient algorithm reduces communication costs in sparse regression.

New bounds for balanced classification improve understanding of imbalanced datasets.

Downsampling can improve generalization in ridgeless linear regression, especially with optimal sketching size.

Efficiently estimates Cox model coefficients without sharing data.

A new method for feature selection robust to noise and design variability.

Study shows how business cycle affects dividend payout based on managerial stock incentives.

Improved MALA method for neural networks uncertainty quantification.

In modern data analysis, random sampling is an efficient and widely-used strategy to overcome the computational difficulties brought by large sample size. In previous studies, researchers conducted random sampling which is according to the input data but independent on the response variable, however the response variab…

Multi-task learning models using Gaussian processes (GP) have been developed and successfully applied in various applications. The main difficulty with this approach is the computational cost of inference using the union of examples from all tasks. Therefore sparse solutions, that avoid using the entire data directly a…

The study finds solar terms significantly impact China's stock market returns and volatility.

The paper explores effective data selection methods for weakly supervised learning.

The paper analyzes how deep models memorize spurious features.

Unified framework for testing deep learning models with concept activation vectors.

Efficiently samples posterior distributions using Langevin dynamics.

For classification problems with significant class imbalance, subsampling can reduce computational costs at the price of inflated variance in estimating model parameters. We propose a method for subsampling efficiently for logistic regression by adjusting the class balance locally in feature space via an accept-reject …

The paper develops methods for causal function estimation and inference with multiway clustered data.

This paper examines cryptocurrency integration with traditional markets, showing how network structure and turbulence influence cross-asset spillovers.

Develops asymptotic analysis for RandNLA sampling estimators in least-squares problems.

OptCS optimizes model selection after conformal inference, controlling FDR and power loss.

Quantum kernels show no advantage in stock return prediction, but differ in stability metrics.

A neural network estimates sampling distributions for hard problems where classical methods fail.

The study finds that firm membership in flagship indices and TCFD endorsement are strong predictors of a wider Disclosure-Performance Gap.