Oracle inequality for sparse neural nets adapts to unknown structure.
problem Sparse deep neural nets in nonparametric regression.
method Gibbs posterior distribution with Metropolis-adjusted Langevin algorithms and mixture of uniform priors.
result Oracle inequality showing adaptation to unknown regularity and structure, achieving minimax-optimal rate of convergence.
Oracle-efficient algorithms reduce combinatorial semi-bandit regret to logarithmic time.
problem Scalability issue in combinatorial semi-bandit problems due to high combinatorial optimization costs.
method Oracle-efficient frameworks that minimize oracle queries while maintaining tight regret guarantees.
result Achieved i l d e O ( T ) ilde{O}(\sqrt{T}) i l d e O ( T ) regret with O ( log log T ) O(\log\log T) O ( log log T ) oracle queries for worst-case linear rewards. Paper introduces structured sparsity estimators for Generalized Linear Models.
problem Estimating structured sparsity in GLMs with debiased estimators.
method Extends Stucky and van de Geer's results to GLMs with structured sparsity.
result Proves oracle inequalities for structured sparsity estimators in GLMs.
A new method for creating derivatives without oracles.
problem Lack of trust in external oracles for derivatives pricing.
method Using Replicating Market Makers (RMMs) to create derivative instruments.
result Demonstrated the feasibility of on-chain expiring options without oracles.
The paper proposes a framework for structured prediction using projection oracles.
problem Structured prediction with improved loss functions.
method A general framework for deriving loss functions using convex sets and projection oracles.
result Projections onto the marginal polytope can make the loss smaller and are computationally efficient.
New adaptive signal denoising method mimics oracle with better statistical properties.
problem Adaptive discrete-time signal denoising with linear oracle structure.
method Minimizes the ℓ 2 \ell_2 ℓ 2 -norm of the estimation residual, proving oracle inequalities for ℓ 2 \ell_2 ℓ 2 -loss. result Improved statistical properties over ℓ ∞ \ell_\infty ℓ ∞ -fit estimators, especially in ℓ 2 \ell_2 ℓ 2 - and pointwise losses. Quantum oracles help identify counterfactuals better than classical ones.
problem Identifying unknown causal parameters in causal models.
method Using quantum oracles to query and identify all causal parameters and counterfactuals.
result Quantum oracles enable identification of all two-way joint counterfactuals and tighter bounds on higher-order counterfactuals.
New algorithm learns FMDP structure while minimizing regret.
problem Regret minimization in FMDPs with unknown structure.
method Optimism in face of uncertainty principle combined with statistical structure learning.
result First algorithm to learn FMDP structure while minimizing regret.
New lower bounds for bilevel optimization with first-order oracles.
problem Complexity of bilevel optimization with first-order oracles.
method Development of hard instances and proof of lower bounds.
result Nontrivial lower bounds for first-order zero-respecting algorithms.
The paper studies the benefits of curriculum learning in linear regression tasks.
problem Theoretical understanding of curriculum learning's benefits in machine learning.
method Theoretical analysis of curriculum learning in structured and unstructured multitask linear regression problems.
result Adaptive learning in the unstructured setting is fundamentally harder than oracle learning, but not in the structured setting.
New algorithms solve stochastic variational inequalities without bounded variance assumption.
problem Solving stochastic variational inequalities without bounded variance assumption.
method Developed algorithms for two classes of problems: monotone and structured nonmonotone VIs.
result Oracle complexity of O(ε^-4) for solving VIs with unbounded domains and possibly unbounded variance.
Paper proposes a method to efficiently estimate structural breaks in cointegrating regressions.
problem Estimating structural breaks in cointegrating regressions is challenging due to inconsistency of group lasso.
method Adaptive group lasso procedure using a first step group lasso estimation of diverging breakpoint candidates to produce weights for a second estimation.
result The adaptive group lasso estimator delivers consistent parameter changes and oracle properties.
Efficient model selection framework for online learning without parameter tuning.
problem Model selection in online learning without predefined parameters.
method Generic meta-algorithm framework for model selection in arbitrary Banach spaces under mild smoothness assumptions.
result First computationally efficient parameter-free algorithms in arbitrary Banach spaces.
Estimates CATEs for structured treatments using a new decomposition method.
problem Estimating conditional average treatment effects for complex data types.
method Generalized Robinson decomposition, isolating causal estimand, arbitrary model plugging, quasi-oracle convergence guarantee.
result Demonstrates superior performance in CATE estimation compared to prior work.
Paper addresses hybrid learning with constrained adversaries, achieving optimal performance.
problem Hybrid learning problem with i.i.d. features and adversarial labels.
method Structured adversarial setting, efficient algorithm with ERM oracle.
result Oracle-efficient algorithm with regret scaling with Rademacher complexity.
Study reconstructs causal graph from latent variables using mixture oracles.
problem Reconstructing causal graphical model from data with latent variables.
method Reduction to mixture oracle to identify latent representations and causal structure.
result Conditions for identifying latent representations and causal model.
Spatial metamodels improve root finding in uncertain oracle responses.
problem Finding roots in noisy, location-dependent oracle responses.
method Propose spatial metamodels to infer oracle distribution and update Bayesian knowledge.
result Spatial PBA algorithm outperforms earlier models in synthetic and real-world problems.
Bayesian-guided method selects optimal design from large candidate pool.
problem Optimizing complex structures with high-fidelity evaluations.
method Bayesian active learning with surrogate modeling.
result Optimal design identified with minimal oracle evaluations.
Develops accelerated fixed-point methods with delayed oracles for scientific computing.
problem Approximating fixed points of nonexpansive operators.
method Combines Nesterov's acceleration and KM iteration with delayed inexact oracles.
result Establishes improved convergence rates for fixed-point approximation.
In this paper, we provide new complexity results for algorithms that learn discrete-variable Bayesian networks from data. Our results apply whenever the learning algorithm uses a scoring criterion that favors the simplest model able to represent the generative distribution exactly. Our results therefore hold whenever t…
Novel framework learns efficient student models from teacher networks.
problem Model capacity gap between teacher and student networks.
method Neural architecture search and oracle knowledge distillation.
result Searched student models often outperform teacher models.
The paper analyzes the complexity of sparse label propagation on networks.
problem Computational complexity of sparse label propagation on network data.
method Characterization of iterations for achieving a prescribed accuracy using a first-order oracle model.
result An upper bound on iterations required for accuracy, showing sharpness for chain structures.
New algorithms sample structured logconcave families with improved efficiency.
problem Sampling structured logconcave families to high accuracy.
method Reduction framework inspired by proximal point methods, combined with restricted Gaussian oracles.
result Improved bounds for sampling structured distributions, matching or surpassing state-of-the-art results.
Study how noisy labels affect semi-supervised learning.
problem Effect of noisy labels on semi-supervised learning performance.
method Proposed an algorithm derived from a continuous relaxation of the Maximum A Posteriori (MAP) estimator for a Degree Corrected Stochastic Block Model (DC-SBM).
result Our approach achieves promising performance even with very noisy labeled data.
New analysis shows Thompson Sampling can work with greedy approximations in combinatorial bandits.
problem Thompson Sampling's theoretical limits with greedy approximations in combinatorial semi-bandits.
method Study with greedy oracle, providing lower and upper bounds on regret.
result First theoretical results showing TS can work with greedy approximations, breaking misconceptions.
A new method learns the optimal pricing map for semiparametric dynamic pricing problems.
problem Optimizing pricing strategies in a semiparametric valuation model with unknown utility and noise.
method Developed a modular policy called ORBIT that uses a scalar pilot index, localizes a benchmark price, and learns a local polynomial approximation of the oracle price map.
result Achieves regret bound of \( \widetilde{O}\big(T^{\frac{2β-1}{4β-3}}+\sqrt{dT}\big) \) for the linear utility model and minimax sharp lower bound.
Paper proposes a new method to minimize submodular functions with fewer calls to simpler oracles.
problem Minimizing the sum of submodular set functions with limited information.
method Introduces a modified convex problem requiring constrained total variation oracles that can be solved with fewer calls to minimization oracles.
result Shows significant reduction in the number of calls to minimization oracles.
New algorithms sample convex bodies using Markov chains and restricted Gaussian oracles.
problem Sampling uniformly from convex bodies efficiently.
method Markov chain Monte Carlo with proximal sampler and restricted Gaussian oracle.
result Efficient implementation of RGO for uniform sampling on convex bodies.
MAMBA learns policies competitive with multiple conflicting oracles.
problem Learning policies from multiple conflicting oracles in reinforcement learning.
method MAMBA uses a gradient estimator in the style of GAE to optimize policies, leveraging demonstrations from multiple weak oracles.
result MAMBA outperforms the state-of-the-art in learning policies competitive with multiple conflicting oracles.
We study the problem of interactively learning a binary classifier using noisy labeling and pairwise comparison oracles, where the comparison oracle answers which one in the given two instances is more likely to be positive. Learning from such oracles has multiple applications where obtaining direct labels is harder bu…
New oracle uses uncertainty for active classification with noisy feedback.
problem Improving query complexity in interactive binary classifier learning.
method Proposes a new pairwise comparison oracle that considers uncertainty and an adaptive labeling algorithm.
result Demonstrates improved performance and efficiency compared to existing methods.
SoQal reduces oracle label requests in active learning by up to 35%.
problem Exploiting unlabelled data in healthcare requires costly oracle labeling.
method Dynamic questioning strategy to minimize oracle label requests.
result SoQal reduces oracle label requests by up to 35%.
Optimal CATE estimation with structured contrast functions using KRR.
problem Estimating CATEs with complex response functions in RKHS.
method Unified two-stage kernel ridge regression method for structured contrast functions.
result Minimax rates governed by contrast function complexity, enabling adaptation.
Paper addresses online alignment of large language models under uncertain preference feedback.
problem Online alignment of large language models with misspecified preference feedback.
method Formulates an oracle-robust objective as a worst-case optimization problem for log-linear policies, and develops projected stochastic composite updates.
result Shows that the robust objective admits an exact closed-form decomposition and achieves O ~ ( ε − 2 ) \widetilde{O}(\varepsilon^{-2}) O ( ε − 2 ) oracle complexity. New model shows online and statistical learning are computationally equivalent with optimization oracle.
problem Online learning in non-convex games with adversarial settings.
method Strengthening the oracle model to make online and statistical learning computationally equivalent.
result Efficient computation of non-convex game equilibria, including GANs, with optimization oracle.
Combines RL and imitation learning with oracle to optimize policies.
problem Optimizing policies in complex environments with limited oracle accuracy.
method Truncated Horizon Policy Search (THOR) using reward shaping.
result THOR achieves superior performance compared to RL and IL baselines.
Study improves CI tests for relational data to robustly discover causal structures.
problem Learning causal relationships from relational data.
method Conduct CI tests against relational data to robustly recover causal structure.
result Effective approach demonstrated through experiments.
New oracles improve stochastic optimization with noisy or biased measurements.
problem Optimizing functions with noisy or biased measurements.
method Introduced biased gradient oracles for stochastic optimization, analyzed RSG and SGD algorithms with these oracles.
result Derived non-asymptotic bounds for convergence rates of algorithms with biased gradient oracles.
Algorithm solves online binary classification and infinite games using ERM oracle.
problem Online learning and solving infinite games with computationally inefficient oracles.
method Proposes an algorithm relying solely on ERM oracle calls for online binary classification and nonparametric games.
result Achieves finite and sublinearly growing regret in various settings.
New bounds on complexity for finding near-stationary points in stochastic convex optimization.
problem Finding near-stationary points in stochastic convex optimization.
method Joint analysis of local stochastic oracle and global oracle models; extensions of recursive regularization technique.
result Logarithmic dependence on smoothness in global oracle model for finding near-stationary points.
Develops Frank-Wolfe Augmented Lagrangian for convex optimization.
problem Minimizing functions over intersections of convex sets.
method Frank-Wolfe Augmented Lagrangian (FW-AL) method.
result Sublinear convergence rate for general convex compact sets, linear for polytopes.
Paper studies how to cluster data with a weak oracle, reducing the number of queries needed.
problem Cluster data with limited oracle answers.
method Proposes algorithms for semi-supervised active clustering with weak oracles.
result Shows that a small number of queries is sufficient for effective clustering.
Representation learning systems typically rely on massive amounts of labeled data in order to be trained to high accuracy. Recently, high-dimensional parametric models like neural networks have succeeded in building rich representations using either compressive, reconstructive or supervised criteria. However, the seman…
Optimizes bilevel empirical risk minimization with improved oracle calls.
problem Optimizing bilevel empirical risk minimization problems.
method Proposes a bilevel extension of the SARAH algorithm.
result Demonstrates improved oracle calls to achieve stationarity.
Max-min margin Markov networks improve consistency in structured prediction.
problem Statistical inconsistency in max-margin methods for structured prediction.
method Defining a max-min margin formulation to overcome statistical inconsistency.
result Proves consistency and provides an explicit algorithm with finite sample generalization bounds.
Paper proposes a new method to stabilize noisy gradient algorithms.
problem Stochastic-gradient Langevin algorithms can introduce bias when taming denominators depend on stochastic-gradient realizations.
method Proposes a structure-preserving framework for designing tamed denominators that avoid unnecessary taming and maintain the stabilizing effect of taming.
result The method avoids stationary bias and explains the stationary error split into bias and remaining error.
New algorithm learns efficiently with a simple 'yes/no' oracle.
problem Can efficient learning be achieved with a simpler oracle than ERM?
method Developed an oracle that returns 'yes' or 'no' for realizable datasets.
result Learnability is possible with a polynomial price in VC dimension.
New method recovers clusters in non-convex finite metric spaces with oracle queries.
problem Exact recovery of clusters in non-convex finite metric spaces.
method Introducing ( β , γ ) (β,γ) ( β , γ ) -convexity and a deterministic algorithm using oracle queries. result Clusters can be recovered using O ( k 2 log n + k 2 ( 6 / β γ ) d e n s ( X ) ) O(k^2 \log n + k^2 (6/βγ)^{dens(X)}) O ( k 2 log n + k 2 ( 6/ β γ ) d e n s ( X ) ) same-cluster queries.