arXiv research

Automatically tunes distance threshold in metric learning.

Sharp threshold found for Frechet mean of inhomogeneous graphs.

Proposes fair classification method using Wasserstein-1 distances.

Locally adaptive clustering for tree delineation.

New algorithm identifies good arms with fewer samples when thresholds are close.

Revisits fuzzy neural networks with generalized Hamming distance, simplifying BN and ReLU.

New method identifies extreme risk propagation in financial networks.

Study examines stock market connections before, during, and after the 2008 financial crisis.

Optimal stopping strategy for a Lévy process near its supremum.

Optimizes risk assessment tools using mixed-integer programming.

The study bounds the stability of Gaussian mixtures under small perturbations.

Paper introduces WWAggr for ensemble CPD, improving accuracy and decision threshold selection.

Study examines financial market structure changes during the COVID-19 crash using a novel MI approach.

PCA helps identify key dimensions for multivariate extremes.

The paper identifies clusters in the World Trade Network using communicability distances.

Transforms distance-based outlier scores into interpretable probabilistic estimates.

Optimal threshold resetting reduces search time for multiple diffusive searchers.

LOT framework speeds up event distance computation in collider physics.

For a certain class of distributions, we prove that the linear programming relaxation of $k$-medoids clustering---a variant of $k$-means clustering where means are replaced by exemplars from within the dataset---distinguishes points drawn from nonoverlapping balls with high probability once the number of points drawn a…

Develops a hypothesis testing framework for generalized Thurstone models.

New method learns local metrics for k-NN classification using sample similarity.

New adaptive tests improve statistical dependence detection.

We improve deep threshold networks' memorization capacity exponentially.

Efficient method classifies locally stationary time series based on second-order characteristics.

The study uses persistent homology to determine when Voronoi interpolation should stop.

In this paper we continue our study of the Laplacian on manifolds with axial analytic asymptotically cylindrical ends initiated in~arXiv:1003.2538. By using the complex scaling method and the Phragmén-Lindelöf principle we prove exponential decay of the eigenfunctions corresponding to the non-threshold eigenvalues of t…

New algorithm achieves strong consistency in binary non-uniform hypergraph classification.

This paper shows how to estimate distances in latent space of random graphs using entropic OT.

Develops a privacy-preserving ABC method using SVT.

Support Vector Data Description (SVDD) is a machine learning technique used for single class classification and outlier detection. SVDD based K-chart was first introduced by Sun and Tsung for monitoring multivariate processes when underlying distribution of process parameters or quality characteristics depart from Norm…

Using data from 92 indices of stock exchanges worldwide, I analize the cluster formation and evolution from 2007 to 2010, which includes the Subprime Mortgage Crisis of 2008, using asset graphs based on distance thresholds. I also study the survivability of connections and of clusters through time and the influence of …

This paper finds the noise threshold for learning Gaussian mixture models equals channel capacity.

We consider the problem of Gaussian mixture clustering in the high-dimensional limit where the data consists of $m$ points in $n$ dimensions, $n,m \rightarrow \infty$ and $α= m/n$ stays finite. Using exact but non-rigorous methods from statistical physics, we determine the critical value of $α$ and the distance between…

Adaptive sampling theory has shown that, with proper assumptions on the signal class, algorithms exist to reconstruct a signal in $\mathbb{R}^{d}$ with an optimal number of samples. We generalize this problem to the case of spatial signals, where the sampling cost is a function of both the number of samples taken and t…

This work learns shared word embeddings for acoustic and phonetic sequences.

New method for optimal transport with missing data, debiased and efficient.

Subspace clustering refers to the problem of clustering high-dimensional data points into a union of low-dimensional linear subspaces, where the number of subspaces, their dimensions and orientations are all unknown. In this paper, we propose a variation of the recently introduced thresholding-based subspace clustering…

Paper uses deep learning for accurate, monotonic cardinality estimation.

A novel approach selects EEGs for better brain disease diagnosis.

This work bounds the run-time of nonconvex optimization with early stopping.

Study on communication delays in decentralized learning networks.

The problem of clustering noisy and incompletely observed high-dimensional data points into a union of low-dimensional subspaces and a set of outliers is considered. The number of subspaces, their dimensions, and their orientations are assumed unknown. We propose a simple low-complexity subspace clustering algorithm, w…

A new algorithm of the analysis of correlation among economy time series is proposed. The algorithm is based on the power law classification scheme (PLCS) followed by the analysis of the network on the percolation threshold (NPT). The algorithm was applied to the analysis of correlations among GDP per capita time serie…

This paper deals with prediction of anopheles number, the main vector of malaria risk, using environmental and climate variables. The variables selection is based on an automatic machine learning method using regression trees, and random forests combined with stratified two levels cross validation. The minimum threshol…

The paper improves conformal prediction by analyzing the beta law of conditional coverage.

Optimal iterative thresholding algorithms improve upon hard and soft thresholding.

BDMBC clusters data with varying densities using a new PLLS measure.

Proposes COLA, a communication-efficient algorithm for decentralized optimization.