Estimates manifold dimension from random samples.
problem Estimating the dimension of a manifold from random samples.
method Explicit theoretical and heuristic bounds for data set size.
result Data set needs to be sufficiently large for accurate dimension estimation.
Paper proposes a new confidence dimension to measure DNN generalization.
problem Measuring the generalization ability of deep neural networks is challenging.
method Introduces confidence dimension (CD) based on Hoeffding's inequality and VC-dimension.
result CD provides a feasible framework to calculate the upper bound of generalization.
The paper improves confidence set construction for statistical inference.
problem Constructing reliable confidence sets in statistical inference.
method Establishes a finite-sample bound using effective dimension and generalized self-concordance.
result Developed a confidence set adapted to optimization landscapes.
Paper develops methods for PCA inference with missing data and heteroskedastic noise.
problem Constructing confidence regions for PCA in high dimensions with missing data and heteroskedastic noise.
method Proposes HeteroPCA and develops non-asymptotic distributional guarantees for valid inference.
result Valid inference on principal subspace and spiked covariance matrix with missing data.
We consider the setting of linear regression in high dimension. We focus on the problem of constructing adaptive and honest confidence sets for the sparse parameter θ, i.e. we want to construct a confidence set for theta that contains theta with high probability, and that is as small as possible. The l_2 diameter of a …
We develop an efficient algorithm to find confidence ellipsoids with volume guarantees in high dimensions.
problem Finding robust confidence ellipsoids in high-dimensional data.
method Polynomial time algorithm using primal-dual structure and geometric Brascamp-Lieb inequality.
result Algorithm finds ellipsoids within a O ( β ) γ d O(β)^{γd} O ( β ) γ d volume factor of best β β β -conditioned ellipsoid. We develop time-uniform confidence spheres for estimating means of random vectors.
problem Sequential mean estimation in high-dimensional spaces.
method Derive time-uniform confidence sphere sequences (CSSs) for various types of random vectors.
result Optimal CSSs for log-concave, sub-Gaussian, and sub- ψ ψ ψ random vectors. The unified approach of Feldman and Cousins allows for exact statistical inference of small signals that commonly arise in high energy physics. It has gained widespread use, for instance, in measurements of neutrino oscillation parameters in long-baseline experiments. However, the approach relies on the Neyman construc…
Develops methods for estimating and providing confidence bands in sparse high-dimensional additive models.
problem Estimating and providing reliable confidence bands for nonparametric components in high-dimensional additive models.
method Integrates sieve estimation into a high-dimensional Z-estimation framework, employing a multiplier bootstrap procedure.
result Constructs uniformly valid confidence bands for the target component f 1 f_1 f 1 in sparse high-dimensional additive models. A new differentiable UCB algorithm for linear bandits learns adaptive confidence bounds.
problem Inability of UCB to strike optimal exploration-exploitation due to confidence bounds.
method Proposes a differentiable linear bandit algorithm and a gradient estimator for learning adaptive confidence bounds.
result Achieves a i l d e O ( β ^ d T ) ilde{\mathcal{O}}(\hatβ\sqrt{dT}) i l d e O ( β ^ d T ) upper bound of T T T -round regret. Algorithm finds small confidence sets for arbitrary distributions.
problem Learning high-density regions in arbitrary distributions.
method Competitive with sets from a concept class with bounded VC-dimension.
result Algorithm finds a confidence set with volume exp ( i l d e O ( d 1 / 2 ) ) \exp( ilde{O}(d^{1/2})) exp ( i l d e O ( d 1/2 )) competitive with optimal ball. Novel confidence sets improve linear bandit performance by adapting to unknown noise levels.
problem Adapting to unknown noise levels in sequential decision-making.
method Proposed semi-adaptive and variance-adaptive confidence sets.
result Improved regret bounds and better performance in Bayesian optimization tasks.
Meta-learned confidence improves few-shot learning accuracy.
problem Improving accuracy in few-shot learning with unreliable model confidence.
method Meta-learning confidence weights for query samples to improve transductive inference performance.
result Meta-learned confidence leads to new state-of-the-art results on benchmark datasets.
Paper certifies intersection of minimum-volume confidence sets for multinomial outcomes.
problem Certifying intersection of minimum-volume confidence sets for multinomial outcomes.
method Exploits likelihood ordering to induce halfspace constraints, enabling adaptive geometric partitioning and computable bounds on p-values.
result Efficient and provably sound algorithm for certifying intersection, disjointness, or indeterminate result.
Improved online confidence bounds for multinomial logistic models in bandits.
problem Achieving optimal regret in multinomial logistic bandits with bounded parameters and outcomes.
method Deriving an improved online confidence bound and proposing OFU-MNL++ and OFU-MN 2 ^2 2 L algorithms. result Achieved variance-dependent optimal regret for MNL bandits.
Prove non-asymptotic bounds for minimal risk in statistical learning
problem Estimating minimal risk in statistical learning
method Using concentration inequalities
result Non-asymptotic bounds for minimal risk
The paper develops methods for constructing confidence regions for regression functions in binary classification.
problem Building distribution-free confidence regions for regression functions in binary classification.
method Resampling test and empirical risk minimization approach for model classes with finite pseudo-dimensions and inverse Lipschitz parameterizations.
result Strong uniform consistency and exponential probably approximately correct bounds on the L 2 L_2 L 2 sizes of the regions. Kernel balancing weights are generalized as KRRR, providing better confidence intervals for treatment effects.
problem Lack of generalization error, correct feature specification, and limited to average effects.
method Interpreting kernel balancing weights as KRRR, relaxing feature specification, and extending Gaussian approximation.
result KRRR provides strong generalization properties and justifies confidence sets for causal functions.
New algorithm tackles non-stationary reinforcement learning with general function approximation.
problem Understanding non-stationary MDPs with function approximation.
method Dynamic Bellman Eluder (DBE) dimension for complexity, sliding window mechanism, confidence set design.
result Upper bound on dynamic regret for proposed SW-OPEA algorithm.
Paper proposes sparse classification method for high-dimensional data.
problem Sparse classification in high-dimensional data with positive-confidence samples.
method Developed a novel sparse-penalization framework using L1, SCAD, and MCP penalties for convex and non-convex shrinkage.
result Proved near minimax-optimal sparse recovery rates under Restricted Strong Convexity condition.
Study optimal policy regret in partially observable Markov games with adaptive opponents.
problem Optimal sequential decision-making in partially observable environments against strategic, adaptive opponents.
method An epoch-based optimistic maximum-likelihood algorithm that selects one policy per epoch using confidence sets built cumulatively from past data.
result Achieves i l d e O ( T ) ilde{O}(\sqrt{T}) i l d e O ( T ) policy regret for fixed problem parameters, with explicit dependence on horizon, adversary memory, confidence radius, and aggregate Eluder dimension. Thompson Sampling with bilateral uncertainty improves performance in Bayesian Optimization.
problem Twin difficulties of modeling and searching complex functions in high dimensions.
method Exploiting conditional independence, Thompson Sampling respecting bilateral uncertainty (BU).
result Thompson Sampling with BU is more effective than the additive approximation in small budgets.
New protocol makes neural MI estimators reliable in high-dimensional data.
problem Accurate estimation of mutual information in high-dimensional, undersampled data.
method Developed a practical protocol for neural MI estimators, incorporating statistical consistency checks, bias correction, and confidence intervals.
result Neural MI estimators can be made reliable when dependencies admit a low-dimensional latent representation.
Efficient local planning with linear approximations for agents with limited simulator access.
problem Planning with limited simulator access in reinforcement learning.
method Confident Monte Carlo Least Square Policy Iteration (Confident MC-LSPI) and Politex (Confident MC-Politex) algorithms.
result The algorithms can learn the optimal policy with local simulator access, even for linear Q-functions.
Hypothesis testing in the linear regression model is a fundamental statistical problem. We consider linear regression in the high-dimensional regime where the number of parameters exceeds the number of samples ( p > n p> n p > n ). In order to make informative inference, we assume that the model is approximately sparse, that is th…
New methods for estimating and inferring nonparametric structural functions and elasticities.
problem Estimating and inferring nonparametric structural functions and their derivatives.
method Data-driven sieve dimension choice and uniform confidence bands construction.
result Optimal estimation and inference procedures with minimax rates of convergence.
We consider the problem of providing nonparametric confidence guarantees for undirected graphs under weak assumptions. In particular, we do not assume sparsity, incoherence or Normality. We allow the dimension D D D to increase with the sample size n n n . First, we prove lower bounds that show that if we want accurate infe…
Unified technique for sequential estimation of convex divergences.
problem Estimating convex divergences between distributions.
method Martingale methods and maximal inequalities for reverse submartingales.
result Valid time-uniform confidence sequences for arbitrary stopping times.
New flexible confidence sequences for robust statistical inference.
problem Creating robust statistical inference methods that work under mild assumptions.
method Proposed a new class of asymptotic time-uniform confidence sequences.
result Sharp asymptotic time-uniform confidence sequences achieved under mild assumptions.
New tighter confidence bounds for sequential kernel regression.
problem Quantifying uncertainty in sequential learning algorithms.
method Martingale tail inequalities and conic programming.
result New confidence bounds are tighter than existing ones.
The paper extends confidence sequences for infinite variance data.
problem Addressing confidence sequences for distributions with infinite variance.
method Establishing lower bounds and deriving tight confidence sequences for relaxed bounded p t h p^{th} p t h -moment distributions. result Derived confidence sequences are tighter than those using Dubins-Savage inequality.
Paper improves Lasso for S&P500 index tracking with post-selection inference.
problem Index tracking for S&P500 with many applications.
method Used Lasso for dimension reduction and post-selection inference.
result Lasso method for S&P500 index tracking shows high performance.
Improves binary classification from positive data with skewed confidence.
problem Skewed confidence in positive data affects the performance of Pconf classifiers.
method Parameterized model of skewed confidence and hyperparameter selection.
result Proposed method effectively cancels out the negative impact of skewed confidence.
Improved algorithms for stochastic linear bandits using tighter confidence sequences.
problem Stochastic linear bandits with improved worst-case regret guarantees.
method Novel tail bound for adaptive martingale mixtures to construct tighter confidence sequences.
result Linear bandit algorithm achieves competitive worst-case regret.
The paper develops optimal confidence regions for categorical data.
problem Constructing tight confidence regions for categorical data.
method Develops new theory for minimum average volume confidence regions.
result Shows optimality of the regions for categorical data and its implications for machine learning.
Confidence intervals are a popular way to visualize and analyze data distributions. Unlike p-values, they can convey information both about statistical significance as well as effect size. However, very little work exists on applying confidence intervals to multivariate data. In this paper we define confidence interval…
Paper presents robust confidence sequences for means with known moment bounds and arbitrary corruption.
problem Tackles robustness to outliers and adversarial corruptions in mean estimation.
method Designs new robust exponential supermartingales to create confidence sequences.
result Achieves optimal width and shows smaller margin of error compared to fixed-time robust methods.
This paper studies the geometry of minimum-volume confidence sets for multinomial parameters.
problem Determining if minimum-volume confidence sets for multinomial outcomes are disjoint.
method Enumerating and covering the continuous regions of the exact p-value function to study the geometry of minimum-volume confidence sets.
result The geometry of minimum-volume confidence sets for multinomial parameters is studied, providing insights into their structure and properties.
Study contextual bandits with stage-wise constraints, proving regret bounds and extending results.
problem Contextual bandits with stage-wise constraints in high probability and expectation settings.
method Upper-confidence bound algorithms for linear and non-linear reward/cost functions, extending to multiple constraints.
result Regret bounds for various settings, including non-linear reward/cost functions.
CoinDICE estimates confidence intervals for unknown behavior policies in reinforcement learning.
problem Estimating value of a target policy using only behavior policy data.
method Function space embedding, generalized empirical likelihood method, Lagrangian optimization.
result Valid confidence intervals with tighter and more accurate estimates than existing methods.
The paper investigates how dataset quality and heterogeneity affect model confidence in machine learning.
problem Understanding how dataset quality and heterogeneity impact model confidence in machine learning.
method The study uses theoretical explanations and experimental demonstrations to investigate the effects of dataset size, label noise, and class heterogeneity on model confidence.
result Label noise reduces model confidence, while reduced dataset size increases it, and class heterogeneity leads to inconsistent confidence across classes.
A new concept of confidence in learning is defined and analyzed.
problem Understanding and quantifying trust in learning processes.
method Formal axioms, continuum measures, vector fields, loss functions.
result Confidence can be represented and optimized in learning.
The paper shows over-confidence in models isn't just due to over-parametrization.
problem Over-confidence in machine learning models, especially in binary classification.
method Theoretical analysis of logistic regression and other binary classification problems.
result Logistic regression is inherently over-confident in certain settings, but over-confidence is not always the case.
The paper proposes a method for constructing confidence sets that adapt to the cardinality of the smallest component of a mean vector.
problem Forming confidence sets for the smallest component of an unknown mean vector.
method Sample splitting and self-normalization approach to test each component for being the smallest, maintaining validity regardless of d d d and n n n . result The proposed tests achieve the local minimax separation rate and robust to heavy-tailed distributions.
Algorithm constructs confidence sets for deep neural networks with PAC guarantees.
problem Ensuring reliable predictions for deep neural networks with high confidence.
method Combines calibrated prediction and learning theory bounds.
result Constructs PAC confidence sets for various deep models.
Paper tackles dynamic assortment with dual contexts, improving revenue in e-commerce.
problem Maximizing revenue in e-commerce with personalized recommendations from vast catalogs.
method Low-rank dynamic assortment model and upper confidence bound approach.
result Regret bound of i l d e O ( ( d 1 + d 2 ) r T ) ilde{O}((d_1+d_2)r\sqrt{T}) i l d e O (( d 1 + d 2 ) r T ) for dynamic assortment problem. New confidence intervals improve treatment effect estimation in randomized experiments.
problem Improving confidence intervals for treatment effects in randomized experiments.
method Systematic exploitation of negative dependence or variance adaptivity.
result Achieved nonasymptotic confidence intervals with the same effective sample size as asymptotic ones.
Develops a method to find costly high-confidence errors in black box models.
problem Finding rare high-confidence errors missed by random sampling.
method Adversarial perturbation-guided search technique to find errors at rates greater than expected given model confidence.
result Our Adversarial Distance search discovers high-confidence errors at a rate greater than expected given model confidence.