archive
Every paper Pith has read. Search by title, abstract, or pith.
883 papers in math.ST · page 12
-
Any species sampling process has an exact finite-mixture form
Exact two-stage finite-mixture representations for species sampling processes
-
Duality yields exact inference for Poisson-Dirichlet hidden Markov models
Exact inference via quasi-conjugacy in two-parameter Poisson-Dirichlet hidden Markov models
-
Duality gives exact closed-form inference for Poisson-Dirichlet HMMs
Exact inference via quasi-conjugacy in two-parameter Poisson-Dirichlet hidden Markov models
-
Singular fluctuation equals specific heat in Bayesian models
Singular Fluctuation as Specific Heat in Bayesian Learning
-
Gaussian transport barycenter yields closed-form invariant features
Invariant Feature Extraction Through Conditional Independence and the Optimal Transport Barycenter Problem: the Gaussian case
-
Linear covariate link yields parametric deconvolution rate
Deconvolution in unlinked linear models
-
Poly-time algo approximates beta-conditioned ellipsoids
Learning Confidence Ellipsoids and Applications to Robust Subspace Recovery
-
This paper applies the feature space decomposition method to derive matching upper and…
Sharp convergence rates for Spectral methods via the feature space decomposition method
-
Survival models gain nothing from overparametrization
Understanding Overparametrization in Survival Models through Interpolation
-
Sub-Gaussian mixture bounds regret by ln ln V_T almost surely
Eventually LIL Regret: Almost Sure $\ln\ln T$ Regret for a sub-Gaussian Mixture on Unbounded Data
-
Causal effects defined on events recover standard treatment measures
A fine-grained look at causal effects in causal spaces
-
Subsampling supplies asymptotic confidence bounds for persistence diagrams
Subsampling Confidence Bound for Persistent Diagram via Time-delay Embedding
-
Gaussians yield directed information estimates with O(log N/sqrt(N)) bounds
Non-Asymptotic Error Bounds for Causally Conditioned Directed Information Rates of Gaussian Sequences
-
Sequential method builds bounds on moments of distributions
A novel sequential method for building upper and lower bounds of moments of distributions
-
Quantum time series reduce to classical geometric regression
Asymptotic inference in a stationary quantum time series
-
Adaptive Lasso needs adjusted degrees of freedom count
Degrees of Freedom in Penalized Regression: Model Selection with Adaptive Penalties
-
Finite-sample ATE intervals via regression adjustment even when p exceeds n
Design-based finite-sample analysis for regression adjustment
-
Constrained graph fused lasso matches unrestricted asymptotics
Asymptotic Distribution of Constrained Nearly-Isotonic Graph Fused Lasso
-
Density estimators on circle and sphere hit optimal rates with direct computation
Rate-optimal and computationally efficient nonparametric estimation on the circle and the sphere
-
Joint Catoni equations match oracle rates under heavy tails
Tuning free Catoni type joint robust estimation
-
Sellers achieve optimal regret in price wars without coordination
Online Price Competition under Generalized Linear Demands
-
Model merges Hawkes and autoregressive dynamics for mixed-scale data
Hawkes autoregressive processes: a new model for multiscale and heterogeneous processes
-
Cauchy methods fix size distortions in predictive regressions
Robust Cauchy-Based Methods for Predictive Regressions
-
Auxiliary covariates lower variance bound for treatment effects
Semi-Supervised Treatment Effect Estimation with Unlabeled Covariates for Prediction-Powered Causal Inference
-
Heavy hitters recovered with exp(sqrt(d)) Fourier coefficients
Model-agnostic super-resolution in high dimensions
-
Behrens-Fisher statistic has exact hypergeometric null density
A unified approach to the Behrens-Fisher problem
-
Vector self-normalized bounds extend past sub-Gaussian tails
Vector-valued self-normalized concentration inequalities beyond sub-Gaussianity
-
MLE for heavy-tailed Pearson VII location is consistent
Asymptotics of the maximum likelihood estimator of the location parameter of Pearson Type VII distribution
-
Critical step size adds stochastic correction to high-dim SGD
Limit Theorems for Stochastic Gradient Descent in High-Dimensional Single-Layer Networks
-
Kim-Milman flow maps stable under Fisher info changes in targets
Stability of the Kim--Milman flow map
-
Search direction yields sublinear regret in online bilevel optimization
Stochastic Regret Guarantees for Online Zeroth- and First-Order Bilevel Optimization
-
Oscillating activation raises perceptron capacity classically
Pseudo quantum advantages in perceptron storage capacity
-
Boundary geometry governs rates for treatment effect curves
Estimation and Inference in Boundary Discontinuity Designs: Distance-Based Methods
-
Transformers hit optimal rates despite test shifts
Optimal In-context Adaptivity and Distributional Robustness of Transformers
-
Stability condition gives CLT for ATE estimators in adaptive trials
Design Stability in Adaptive Experiments: Implications for Treatment Effect Estimation
-
Geometric invariance of normals yields Itô formula and Black-Scholes
Topics in Probability, Parametric Estimation and Stochastic Calculus
-
Inflating min-norm solution lowers test error in anisotropic regression
Shrinkage to Infinity: Reducing Test Error by Inflating the Minimum Norm Interpolator in Linear Models
-
Inference framework tests variable importance in treatment effects
Inference on Variable Importance for Treatment Effect Heterogeneity: Shapley Values and Beyond
-
Higher-order Langevin cuts parallel processors for sampling
Fast and Efficient Parallel Sampling Using Higher Order Langevin Dynamics
-
Centered MA innovation matches digamma-link DARMA to first order
Centered-Innovation MA for Bayesian Dirichlet ARMA: Theoretical Equivalence and an Application to Bank-Asset Shares
-
Truncation enables optimal high-dimensional volatility estimation
A robust and scalable estimation for high-dimensional volatility models
-
Bayesian batch and online learning coincide
Batch learning equals online learning in Bayesian supervised learning
-
-
Debiased kernels extend volatility estimation to jump activity 20/11
Debiased Kernel Estimation of Spot Volatility in the Presence of Infinite Variation Jumps
-
Local graph structure sets Wasserstein estimation rates
Fast Wasserstein rates for estimating probability distributions of probabilistic graphical models
-
The paper develops a quasi-maximum likelihood estimator for the unknown drift parameter…
Drift estimation for rough processes under small noise asymptotic : QMLE approach
-
Ginibre product max eigenvalue converges to alpha-dependent limit
Precise convergence rate of spectral radius of product of complex Ginibre
-
Cumulants bounded by moments with (n-1)!/rho^n growth
Explicit Universal Bounds for Cumulants via Moments
-
Bayesian posteriors stabilize LLM rankings better than Pass@k
Don't Pass@k: A Bayesian Framework for Large Language Model Evaluation
-
Method keeps classification error below alpha with sqrt(T) extra tests
The Good, the Bad, and the Sampled: a No-Regret Approach to Safe Online Classification