arXiv research

Generates hard graph instances for algorithm analysis and benchmarking.

A new method generates MAX-2-SAT instances with adjustable hardness.

HardVis helps visually manage imbalanced data by sampling hard instances.

SLR tackles sparse linear regression problems, showing hardness for efficient algorithms.

In Machine Learning, ensemble methods have been receiving a great deal of attention. Techniques such as Bagging and Boosting have been successfully applied to a variety of problems. Nevertheless, such techniques are still susceptible to the effects of noise and outliers in the training data. We propose a new method for…

Neural model with parameterized algorithms improves graph CO problem solving.

Explicitly constructed 3XOR instances hard for Sum-of-Squares hierarchy.

Mathematical Reinforcement Learning faces a 'Two-Hump' problem due to sparse rewards and a scarcity of intermediate 'hard-but-solvable' instances.

Study shows challenges in reinforcement learning math problems, proposing enhancements and a hardness measure.

It is well known that Sparse PCA (Sparse Principal Component Analysis) is NP-hard to solve exactly on worst-case instances. What is the complexity of solving Sparse PCA approximately? Our contributions include: 1) a simple and efficient algorithm that achieves an $n^{-1/3}$-approximation; 2) NP-hardness of approximatio…

Quantum circuits are hard to learn on average.

Learning shrinks hard tail, improving inference performance.

New algorithms minimize regret with multiple best arms in large action spaces.

Combinatorial auctions are formulated as frustrated lattice gases on sparse random graphs, allowing the determination of the optimal revenue by methods of statistical physics. Transitions between computationally easy and hard regimes are found and interpreted in terms of the geometric structure of the space of solution…

Paper tackles instance-dependent label noise by approximating it with part-dependent noise.

A new PU classifier PUAL tackles trifurcate data issues.

A new dynamic attention model improves vehicle routing problem solutions.

Pruning of redundant or irrelevant instances of data is a key to every successful solution for pattern recognition. In this paper, we present a novel ranking-selection framework for low-length but highly correlated instances. Instead of working in the low-dimensional instance space, we learn a supervised projection to …

Proposes an IRT-based ensemble method to improve machine learning accuracy.

Improved hardness results for clearing payments in financial networks with CDSs.

We prove that the problem of deciding whether a 2- or 3-dimensional simplicial complex embeds into $\mathbb{R}^3$ is NP-hard. Our construction also shows that deciding whether a 3-manifold with boundary tori admits an $\mathbb{S}^{3}$ filling is NP-hard. The former stands in contrast with the lower dimensional cases wh…

Paper investigates hardness of learning neural networks under manifold hypothesis.

New RF dissimilarity measures improve multi-view learning accuracy.

We investigate the complexity of finding an embedded non-orientable surface of Euler genus $g$ in a triangulated $3$-manifold. This problem occurs both as a natural question in low-dimensional topology, and as a first non-trivial instance of embeddability of complexes into $3$-manifolds. We prove that the problem is NP…

This paper strengthens the computational separation between multimodal and unimodal learning, showing unimodal learning is hard on typical instances.

We present a new anytime algorithm that achieves near-optimal regret for any instance of finite stochastic partial monitoring. In particular, the new algorithm achieves the minimax regret, within logarithmic factors, for both "easy" and "hard" problems. For easy problems, it additionally achieves logarithmic individual…

Private KL distribution estimation improved with instance-optimality.

The restricted isometry property (RIP) for design matrices gives guarantees for optimal recovery in sparse linear models. It is of high interest in compressed sensing and statistical learning. This property is particularly important for computationally efficient recovery methods. As a consequence, even though it is in …

A method learns to solve multilevel combinatorial problems with two players.

Study shows computational hardness can improve adversarial robustness in learning.

New algorithm recovers labels from noisy categorical data.

New research shows how preconditioning can solve sparse linear regression problems efficiently.

Paper estimates optimal classification error with soft labels and calibration.

Differentiable relaxation for inferring partial orders from noisy linear data.

Matching one set of objects to another is a ubiquitous task in machine learning and computer vision that often reduces to some form of the quadratic assignment problem (QAP). The QAP is known to be notoriously hard, both in theory and in practice. Here, we investigate if this difficulty can be mitigated when some addit…

Surveying a new method to predict computational hardness in hypothesis testing.

Proposes an online pool generation method for DCS to improve classifier selection accuracy.

We propose a framework for Semi-Supervised Active Clustering framework (SSAC), where the learner is allowed to interact with a domain expert, asking whether two given instances belong to the same cluster or not. We study the query and computational complexity of clustering in this framework. We consider a setting where…

New method estimates optimal Q-values with better accuracy for specific problems.

Paper establishes lower bounds for finite-sum optimization problems using novel construction methods.

Study shows computational limits for robust classification tasks, leading to cryptographic implications.

Annealed Entropic Allocation improves ranking and selection by mitigating hard switching and improving finite-budget discrimination.

Algorithm finds optimal regularizers for online linear optimization.

We investigate the relation of two fundamental tools in machine learning and signal processing, that is the support vector machine (SVM) for classification, and the Lasso technique used in regression. We show that the resulting optimization problems are equivalent, in the following sense. Given any instance of an $\ell…

This paper proposes Relational Similarity Machines (RSM): a fast, accurate, and flexible relational learning framework for supervised and semi-supervised learning tasks. Despite the importance of relational learning, most existing methods are hard to adapt to different settings, due to issues with efficiency, scalabili…

Optimal sample complexity for learning Plackett-Luce models.

Study shows it's impossible to count communities without finding them.

Optimal transport is #P-hard when components are independent, even with approximate solutions.