A new method corrects for bias in selecting the best candidate.
problem Bias in selecting the best candidate leads to invalid conclusions.
method Zoom correction, flexible for parametric and nonparametric settings.
result Valid inference on the winner is possible, even with selection bias.
Data-driven decision-making often overestimates benefits due to the winner's curse.
problem Accurate policy evaluation in data-driven decision-making.
method Model-based policy evaluation using estimated models from data.
result Model-based methods can produce large, spurious reported benefits even when true effects are zero.
New method optimizes treatment policies to avoid winner's curse.
problem Winner's curse in treatment policy optimization.
method Inference-aware policy optimization.
result Optimizes for both estimated performance and downstream evaluation.
SIREN protocol corrects optimistic winner's scores in LLM evaluation.
problem Optimistic winner's scores in LLM evaluation due to adaptive benchmarking.
method SIREN protocol that freezes post-search shortlist, separates selection and evaluation, and uses bootstrap for uncertainty quantification.
result SIREN provides valid confidence intervals for procedure-performance curves and deployment conclusions.
New algorithm reduces dynamic regret in non-stationary dueling bandits using a weighted Borda score.
problem Designing algorithms with low dynamic regret in non-stationary dueling bandits.
method Introducing a novel weighted Borda score framework to analyze the Condorcet problem and establish improved bounds.
result First optimal and adaptive dynamic regret upper bound of i l d e O ( i l d e L 1 / 3 K 1 / 3 T 2 / 3 ) ilde{O}( ilde{L}^{1/3} K^{1/3} T^{2/3} ) i l d e O ( i l d e L 1/3 K 1/3 T 2/3 ) . New method identifies Condorcet winner in dueling bandits with improved sample complexity.
problem Identifying Condorcet winner in noisy pairwise comparisons.
method Exploits full gap matrix Δ to improve sample complexity.
result Improves sample complexity guarantees by leveraging informative comparisons.
This paper tackles combinatorial pure exploration for dueling bandits, aiming to find the best candidate-position match.
problem Finding the best candidate-position match in a dueling bandit setting.
method The paper adapts combinatorial pure exploration for multi-armed bandits to dueling bandits, considering both Borda winner and Condorcet winner cases. It designs PAC and exact algorithms for Borda winner and a fully polynomial time approximation scheme (FPTAS) for Condorcet winner.
result The paper introduces the first algorithm with polynomial running time per round for identifying the Condorcet winner in CPE-DB.
New method uses geometric properties for better density estimation.
problem Uncertainty quantification in ambiguous tasks.
method Winner-takes-all training with centroidal Voronoi tessellations.
result Improved quantization and density estimation.
Algorithm identifies Copeland winners in dueling bandits with ternary feedback.
problem Identifying Copeland winners in dueling bandits with indifferences.
method Proposed POCOWISTA algorithm with a sample complexity close to lower bound.
result Algorithm shows excellent performance, even for conventional dueling bandits.
A game-theoretic approach to multi-criteria ranking from ordinal data.
problem Ranking objects from ordinal data with multiple criteria.
method Generalizing von Neumann winner to multi-criteria setting using Blackwell's approachability.
result The Blackwell winner can be computed as a convex optimization problem and achieves near-optimal sample complexity.
Study shows big winner stocks significantly impact passive and active investment strategies.
problem Impact of big winner stocks on passive and active investment strategies.
method Numerical and analytical techniques applied to historical stock price data.
result Concentrated portfolios underperform equally weighted indexes due to missing big winner stocks.
This paper reexamines the profitability of loser, winner and contrarian portfolios in the Chinese stock market using monthly data of all stocks traded on the Shanghai Stock Exchange and Shenzhen Stock Exchange covering the period from January 1997 to December 2012. We find evidence of short-term and long-term contraria…
TimeMCL forecasts diverse time series futures using neural networks and WTA loss.
problem Forecasting multiple plausible time series futures.
method Multiple Choice Learning (MCL) framework with Winner-Takes-All (WTA) loss.
result TimeMCL efficiently predicts diverse time series futures at low computational cost.
Stochastic LWTA networks resist adversarial attacks while maintaining accuracy.
problem Adversarial robustness of neural networks.
method Replaced ReLU with stochastic LWTA activations, trained with Variational Bayesian and PGD.
result Stochastic LWTA networks achieve state-of-the-art robustness against adversarial attacks.
Paper proposes a method to predict MOBA game winners with calibrated confidence.
problem Predicting MOBA game winners with noisy data and uncertain noise.
method A novel confidence-calibration method considering data uncertainty.
result Achieves outstanding expected calibration error (ECE) of 0.57%.
aMCL uses annealing to improve hypothesis diversity in ambiguous tasks.
problem Limitations of Winner-takes-all in predicting plausible hypotheses.
method Combines simulated annealing with Multiple Choice Learning (MCL).
result Enhanced exploration of hypothesis space during training.
This case study tests the possibility of prediction for "success" (or "winner") components of four stock & shares market indices in a time period of three years from 02-Jul-2009 to 29-Jun-2012.We compare their performance ain two time frames: initial frame three months at the beginning (02/06/2009-30/09/2009) and the f…
In our model, n n n traders interact with each other and with a central bank; they are taxed on the money they make, some of which is dissipated away by corruption. A generic feature of our model is that the richest trader always wins by 'consuming' all the others: another is the existence of a threshold wealth, below wh…
Sentiment Analysis of microblog feeds has attracted considerable interest in recent times. Most of the current work focuses on tweet sentiment classification. But not much work has been done to explore how reliable the opinions of the mass (crowd wisdom) in social network microblogs such as twitter are in predicting ou…
Public debates are a common platform for presenting and juxtaposing diverging views on important issues. In this work we propose a methodology for tracking how ideas flow between participants throughout a debate. We use this approach in a case study of Oxford-style debates---a competitive format where the winner is det…
DPLTM uses private winners to improve privacy and utility.
problem Balancing privacy and utility in machine learning models.
method DPLTM applies differential privacy to the lottery ticket method.
result DPLTM converges faster and uses less privacy budget.
We consider the problem of evaluating the quality of startup companies. This can be quite challenging due to the rarity of successful startup companies and the complexity of factors which impact such success. In this work we collect data on tens of thousands of startup companies, their performance, the backgrounds of t…
Algorithm minimizes regret in non-stationary dueling bandits with unknown parameters.
problem Minimizing regret in dueling bandits with time-varying preferences.
method Proposes Beat the Winner Reset algorithm and meta-algorithms DETECT and Monitored Dueling Bandits.
result Proves bounds on expected weak and strong regret for non-stationary dueling bandits.
New findings suggest Barron space doesn't defy curse of dimensionality for certain types of smoothness.
problem Understanding the curse of dimensionality in neural networks with different smoothness notions.
method Defined ADZ spaces via Mellin transform to encapsulate nonclassical smoothness, compared to classical smoothness.
result Evidence provided that Barron space doesn't defy curse of dimensionality for certain smoothness types.
We study the K-armed dueling bandit problem, a variation of the standard stochastic bandit problem where the feedback is limited to relative comparisons of a pair of arms. The hardness of recommending Copeland winners, the arms that beat the greatest number of other arms, is characterized by deriving an asymptotic regr…
Unified framework for best arm identification and dueling bandits regret minimization.
problem Best arm identification and dueling bandits regret minimization.
method Tree-Guided Identify-Then-Exploit (TG-ITE) framework.
result Unified approach achieving optimal sample complexity and regret guarantees.
New method avoids curse of dimensionality in structured density estimation.
problem Estimating multivariate density with Markov graph constraints.
method Introduces 'graph resilience' to control sample complexity.
result Avoids curse of dimensionality under Markov conditions.
New algorithms improve dueling bandit performance in multiplayer settings.
problem Challenges in collaborative exploration of non-informative arm pairs in multiplayer dueling bandits.
method Demonstrated Follow Your Leader approach and message-passing fully distributed protocol.
result Multiplayer algorithms outperform single-player benchmarks.
Deep neural nets can estimate regression with dependent data without the curse of dimensionality.
problem Regression with dependent data and structural assumptions on the regression function.
method Deep recurrent neural network estimate under suitable structural assumptions.
result Deep neural nets can circumvent the curse of dimensionality for regression with dependent data.
The curse of dimensionality affects neural network optimization, especially with smooth functions.
problem The curse of dimensionality in neural network optimization.
method Examined through the evolution of the parameter distribution under 2-Wasserstein gradient flow.
result The curse of dimensionality persists in neural network optimization, even with smooth functions.
New method circumvents curse of dimensionality in Laplacian estimation.
problem High-dimensional data challenges spectral clustering and diffusion maps.
method Kernelized Laplacian estimation via reproducing kernel Hilbert space.
result Non-asymptotic statistical rates show improved performance in high dimensions.
Study confirms fractional norms and quasinorms do not help overcome curse of dimensionality.
problem Overcoming the curse of dimensionality in machine learning.
method Systematic testing of fractional norms and quasinorms (p<1) on classification problems.
result Distance concentration behavior is qualitatively the same for all norms and quasinorms as dimensionality increases.
We study the effect of altruism in two simple asset exchange models: the yard sale model (winner gets a random fraction of the poorer player's wealth) and the theft and fraud model (winner gets a random fraction of the loser's wealth). We also introduce in these models the concept of bargaining efficiency, which makes …
We propose a simple change to existing neural network structures for better defending against gradient-based adversarial attacks. Instead of using popular activation functions (such as ReLU), we advocate the use of k-Winners-Take-All (k-WTA) activation, a C0 discontinuous function that purposely invalidates the neural …
New method optimizes expensive simulations for complex systems.
problem Optimizing complex systems with limited expensive experiments.
method Black-box Optimization via Marginal Means (BOMM) approach.
result BOMM improves optimization performance in high dimensions.
AI predicts stock winners with 2.43 Sharpe ratio, but returns are highly concentrated.
problem Predicting stock returns with AI, focusing on identifying top winners.
method Deployed a state-of-the-art LLM to autonomously search the web for stock attractiveness, avoiding look-ahead bias.
result AI can generate alpha by identifying top winners, but returns are highly concentrated.
Improved algorithm for adaptive dueling bandits with near-optimal regret bound.
problem Non-stationary dueling bandits with unknown number of preference changes.
method Elimination-based rescheduling algorithm for adaptive dynamic regret.
result Near-optimal i l d e O ( S e x t t t C W T ) ilde{O}(\sqrt{S^{ exttt{CW}} T}) i l d e O ( S e x ttt C W T ) dynamic regret bound. LoRA-MCL improves language models by generating diverse sentence continuations.
problem Language models struggle with generating diverse, plausible sentence continuations.
method Low-Rank Adaptation combined with Multiple Choice Learning (MCL) to handle ambiguity.
result LoRA-MCL generates high-diversity and relevant outputs in various tasks.
This paper tackles the curse of dimensionality in semi-supervised learning using Laplacian regularization.
problem The curse of dimensionality in semi-supervised learning with Laplacian regularization.
method Statistical analysis and spectral filtering methods using kernel methods.
result The paper provides a method to overcome the curse of dimensionality in semi-supervised learning.
New algorithms for batched dueling bandits with improved regret bounds.
problem Batched dueling bandits with noisy pairwise comparisons.
method Developed algorithms for two settings: Condorcet winner and strong stochastic transitivity.
result Regret bounds match sequential bounds using only a logarithmic number of batches.
Inspired by the advances in biological science, the study of sparse binary projection models has attracted considerable recent research attention. The models project dense input samples into a higher-dimensional space and output sparse binary data representations after the Winner-Take-All competition, subject to the co…
Paper provides finite-sample guarantees for Wasserstein DRO without dimensionality curse.
problem Tackles empirical success of Wasserstein DRO in operations and ML with performance guarantees.
method Develops non-asymptotic framework for analyzing out-of-sample performance and generalization bound.
result First finite-sample guarantee for generic Wasserstein DRO problems without curse of dimensionality.
Smooth DNNs mitigate the curse of dimensionality in uniform convergence for various regression tasks.
problem The curse of dimensionality in uniform convergence of ReLU networks.
method Analysis of smoothly activated deep neural networks (smooth DNNs), establishing pseudo-dimension bounds and non-asymptotic approximation guarantees.
result Smooth DNNs achieve non-asymptotic uniform convergence rates across multiple statistical contexts, mitigating the curse of dimensionality.
HKRR adapts to MIM, overcoming the curse of dimensionality.
problem Understanding when deep networks outperform kernel methods in high dimensions.
method Hyper-kernel ridge regression (HKRR) for multi-index models (MIM).
result HKRR can adaptively learn MIM, overcoming the curse of dimensionality.
New optimization method corrects data-driven optimizer's curse.
problem Over-optimistic evaluation in data-driven optimization.
method Smoothed f f f -Divergence Distributionally Robust Optimization (DRO). result Statistical bound on out-of-sample performance nearly tightest.
SCORE technique reduces BO's high-dimensional search costs.
problem Bayesian optimization's high computational costs in high-dimensional spaces.
method 1D reparametrization trick to maintain linear time complexity.
result Successfully finds global minimum in high-dimensional optimization.
A fast ML method solves complex combinatorial auction problems.
problem Solving winner determination in multi-unit combinatorial auctions.
method Graph Neural Network (GNN) with half-convolution operations for bid-item graph modeling.
result Approaches optimal performance with negligible revenue loss and low complexity.
Input-dependent smoothing mitigates classical issues but suffers from the curse of dimensionality.
problem Certifiably robust classifiers with input-dependent smoothing suffer from the curse of dimensionality.
method Proposed a theoretical and practical framework for input-dependent smoothing under strict restrictions.
result Input-dependent smoothing mitigates some classical issues but is limited by the curse of dimensionality.