SMOTE: Synthetic Minority Over-sampling Technique improves classifier performance on imbalanced datasets by generating synthetic minority samples, boosting AUC by over 0.07 in various experiments.

N. V. Chawla, K. W. Bowyer, L. O. Hall et al.

Proposes a strategy for continuum-armed bandits without knowing the Lipschitz constant, achieving optimal regret bounds.

Sébastien Bubeck, Gilles Stoltz, Jia Yuan Yu

The paper proposes minimax policies for combinatorial prediction games using Bregman projections and potential-based gradient descent to address worst-case minimax regret.

Jean-Yves Audibert, Sebastien Bubeck, Gabor Lugosi

Introduces RFCI, a fast algorithm for high-dimensional causal graph learning with latent and selection variables, ensuring asymptotic correctness.

Diego Colombo, Marloes H. Maathuis, Markus Kalisch et al.

Proposes column subset-based low-rank matrix reconstruction achieving an optimal Frobenius norm approximation ratio of \(\sqrt{rac{r+1}{r-k+1}}\), with algorithms running in O(r n m^ω log m).

Venkatesan Guruswami, Ali Kemal Sinop

Proposes a unified neural network framework leveraging large-scale unlabeled data for multiple NLP tasks, achieving state-of-the-art results without task-specific engineering.

Ronan Collobert, Jason Weston, Leon Bottou et al.

The KL-UCB algorithm outperforms UCB in bounded stochastic bandits, achieving optimal bounds in Bernoulli rewards.

Aurélien Garivier, Olivier Cappé

Using the LePage decomposition, the paper proves that every exp-1-stable point process is a Poisson process of random clusters.

Pascal Maillard

Convex NMF model with l_{1,∞} regularization selects data subset as dictionary, enabling physically meaningful dimensionality reduction.

Ernie Esser, Michael Möller, Stanley Osher et al.

Bayesian optimization efficiently maximizes expensive black-box functions using Gaussian processes and acquisition functions, reducing sample complexity.

Eric Brochu, Vlad M. Cora, Nando de Freitas

Kernel PCA-based nonlinear control system reduction in RKHS, capturing essential dynamics via data-driven Gramian diagonalization.

Jake Bouvrie, Boumediene Hamzi

By analyzing the zero-fluctuation limit, the paper links entropy minimization to Monge-Kantorovich optimal transport via Gamma-convergence and large deviations.

Christian Léonard

Tropp improves analysis of subsampled randomized Hadamard transform, achieving optimal constants for embedding dimension bounds.

Joel A. Tropp

MESA is an open-source 1D stellar evolution code with modular microphysics, adaptive mesh, and parallelism, enabling detailed simulations from low-mass to massive stars.

Bill Paxton, Lars Bildsten, Aaron Dotter et al.

Proposed a dynamic recombination algorithm for high-order cubature on Wiener space, reducing particle explosion.

C. Litterer, T. Lyons

Generalizes Feynman tree theorem to L-loop amplitudes via on-shell phase space integrals, exploiting causality and response functions.

Simon Caron-Huot

Interpolation of scattered data on embedded submanifolds using restricted positive definite kernels with Sobolev error estimates.

Edward Fuselier, Grady Wright

Distributed dual averaging algorithm's convergence rate scales inversely with network spectral gap, enabling efficient large-scale optimization.

John Duchi, Alekh Agarwal, Martin Wainwright

Homophily and contagion are generically confounded in observational social network studies, requiring strong assumptions to distinguish.

Cosma Rohilla Shalizi, Andrew C. Thomas

Optimization under unknown constraints using Gaussian processes and Bayesian learning, applied to healthcare policy.

Robert B. Gramacy, Herbert K. H. Lee