arXiv research

Study identifies current coupons for mortgage-backed securities.

The paper improves boundary detection and density estimation on noisy data.

We propose an efficient method for estimating covariate effects in doubly-stochastic spatial models.

Paper proposes Sinkformers for Transformers with doubly stochastic attention.

Consistency of DSDP proved with exponential convergence.

We propose a unified framework for equity and credit risk modeling, where the default time is a doubly stochastic random time with intensity driven by an underlying affine factor process. This approach allows for flexible interactions between the defaultable stock price, its stochastic volatility and the default intens…

Doubly-stochastic normalization improves robustness to heteroskedastic noise.

Robustly infers manifold density and geometry under high-dimensional noise.

New method reduces variance in complex probabilistic model optimization.

Doubly SGD improves convergence for intractable objective optimization.

Proposes SDRG to adjust missingness in machine learning models.

Characterizes and examines gradient solitons on doubly warped product manifolds.

New method solves doubly-nonconvex composite optimization problems.

A new algorithm solves bilinear saddle-point problems efficiently.

DS-FACTO optimizes factorization machines for large-scale datasets.

Improved nearest neighbors for missing data in latent factor models.

New method improves off-policy critic evaluation in reinforcement learning.

Study explains mortgage burnout using Cox hazard models.

A scalable GPLVM model using stochastic variational inference.

The paper analyzes generalization properties of scalable kernel methods.

Paper optimizes sample selection for top-k arms in stochastic bandits.

FDSKL algorithm trains vertically partitioned data with kernels securely and efficiently.

New estimator for causal effects in large datasets.

New method predicts outcomes even when some factors are not used in models.

Geometric approach for unsupervised word embedding alignment.

DSVNP uses global and local latent variables for improved neural process predictions.

New model improves DNA methylation data analysis.

This paper discusses properties of a Doubly Stochastic Poisson Process (DSPP) where the intensity process belongs to a class of affine diffusions. For any intensity process from this class we derive an analytical expression for probability distribution functions of the corresponding DSPP. A specification of our results…

A new algorithm improves regret bounds for contextual bandits.

Develops a method for causal inference with noisy confounders.

The paper examines conditions for Einstein-like classes in doubly warped product manifolds.

Clustering analysis by nonnegative low-rank approximations has achieved remarkable progress in the past decade. However, most approximation approaches in this direction are still restricted to matrix factorization. We propose a new low-rank learning method to improve the clustering performance, which is beyond matrix f…

The study introduces a new curvature concept for weighted graphs and applies it to warped products.

ADSGD method speeds up model identification in sparse optimization.

Efficiently approximates softmax probabilities for large-scale inference.

We study a doubly reflected backward stochastic differential equation (BSDE) with integrable parameters and the related Dynkin game. When the lower obstacle $L$ and the upper obstacle $U$ of the equation are completely separated, we construct a unique solution of the doubly reflected BSDE by pasting local solutions and…

New algorithms optimize machine learning with both features and observations distributed across a cluster.

π-GNN learns soft permutations for graph representations, improving graph classification and regression.

Accelerates Cox partial-likelihood optimization with improved SGD.

We solve inference for deep Gaussian processes without layer independence.

New algorithms learn graph structures privately, matching best results.

Bayesian Neural Networks built block-by-block with uncertainty estimates.

Solves a Dirichlet problem for flat metrics on Riemann surfaces with boundary.

Proposes a new simulator for complex arrival processes.

The general perception is that kernel methods are not scalable, and neural nets are the methods of choice for nonlinear learning problems. Or have we simply not tried hard enough for kernel methods? Here we propose an approach that scales up kernel methods using a novel concept called "doubly stochastic functional grad…

Enhances DGPs with adaptive RKHS Fourier features for better non-stationary pattern modeling.

S2M optimizes mining for diverse data subpopulations.

A new algorithm speeds up optimal transport for machine learning.