arXiv research

Neural networks struggle with long sequences, but a new method improves their performance.

MACRO meta-algorithm learns from sequentially arriving data without storing all data.

Machine learning improves optimization algorithms in data science.

Study shows effectiveness of offline RL in online RL tasks.

Method estimates dataset utility via minimal program length proxy.

New algorithm achieves optimal regret in average reward MDPs without prior bias information.

Efficient algorithm learns direct causes and effects from data.

A new algorithm improves offline reinforcement learning robustness.

The paper develops a new algorithm for constructing minimax estimators using online learning techniques.

Reverse annealing boosts quantum matrix factorization performance.

Quantum algorithm speeds up Lasso regression by quadratically faster per iteration.

A framework for private prediction sets using conformal prediction and differential privacy.

Previous work in hierarchical reinforcement learning has faced a dilemma: either ignore the values of different possible exit states from a subroutine, thereby risking suboptimal behavior, or represent those values explicitly thereby incurring a possibly large representation cost because exit values refer to nonlocal a…

FPML algorithm reduces regret by limiting the number of arms chosen per round.

The conjugate gradient (CG) method is an efficient iterative method for solving large-scale strongly convex quadratic programming (QP). In this paper we propose some generalized CG (GCG) methods for solving the $\ell_1$-regularized (possibly not strongly) convex QP that terminate at an optimal solution in a finite numb…

This paper explores optimising acquisition functions in Bayesian optimisation.

A new approach for specifying and synthesizing subroutines for optimizing metrics.

DIFF2 improves differential privacy in nonconvex optimization with better utility bounds.

An optimal algorithm maximizes submodular functions online with no-1/2 regret.

New algorithm helps avoid traps in reinforcement learning.

Proposes a new DQN framework using variational inference.

We describe a new optimization scheme for finding high-quality correlation clusterings in planar graphs that uses weighted perfect matching as a subroutine. Our method provides lower-bounds on the energy of the optimal correlation clustering that are typically fast to compute and tight in practice. We demonstrate our a…

Study on encoding neural architectures for NAS, showing impact on performance.

Measures compositionality in machine learning representations.

A new method learns DAGs from data using permutation optimization.

New algorithm learns POMDPs without computational oracles.

New algorithm calibrates forecasts in high dimensions with minimal regret.

Paper proposes an algorithm for lifelong learning with shared structure.

We present a novel certified and complete algorithm to compute arrangements of real planar algebraic curves. It provides a geometric-topological analysis of the decomposition of the plane induced by a finite number of algebraic curves in terms of a cylindrical algebraic decomposition. From a high-level perspective, the…

We study the problem of off-policy value evaluation in reinforcement learning (RL), where one aims to estimate the value of a new policy based on data collected by a different policy. This problem is often a critical step when applying RL in real-world problems. Despite its importance, existing general methods either h…

A parallel Fortran framework for neural networks and deep learning.

A quantum circuit designed for efficient statistical model preparation and training.

New algorithm learns Bayesian networks robustly in nearly-linear time.

We present a novel hybrid algorithm for Bayesian network structure learning, called Hybrid HPC (H2PC). It first reconstructs the skeleton of a Bayesian network and then performs a Bayesian-scoring greedy hill-climbing search to orient the edges. It is based on a subroutine called HPC, that combines ideas from increment…

Solves batch policy learning with constraints using flexible meta-algorithm and OPE.

This paper studies the evaluation of policies that recommend an ordered set of items (e.g., a ranking) based on some context---a common scenario in web search, ads, and recommendation. We build on techniques from combinatorial bandits to introduce a new practical estimator that uses logged data to estimate a policy's p…

Structured sparsity is an important modeling tool that expands the applicability of convex formulations for data analysis, however it also creates significant challenges for efficient algorithm design. In this paper we investigate the generalized conditional gradient (GCG) algorithm for solving structured sparse optimi…

Paper develops fast method for computing optimal transport.

New LNS framework improves integer program solving.

Grinch efficiently clusters large datasets with complex structures.

This thesis explores GNNs, categorizing them into local and global approaches.

cpSGD reduces communication and maintains privacy in distributed learning.

New framework improves fairness in small data settings.

We reduce a broad class of machine learning problems, usually addressed by EM or sampling, to the problem of finding the $k$ extremal rays spanning the conical hull of a data point set. These $k$ "anchors" lead to a global solution and a more interpretable model that can even outperform EM and sampling on generalizatio…

We study modeling and inference with the Elliptical Gamma Distribution (EGD). We consider maximum likelihood (ML) estimation for EGD scatter matrices, a task for which we develop new fixed-point algorithms. Our algorithms are efficient and converge to global optima despite nonconvexity. Moreover, they turn out to be mu…

Discovering the latent structure from many observed variables is an important yet challenging learning task. Existing approaches for discovering latent structures often require the unknown number of hidden states as an input. In this paper, we propose a quartet based approach which is \emph{agnostic} to this number. Th…

A new unsupervised method for molecule representations.

Quantum algorithms improve perceptron learning efficiency.