arXiv research

Deep learning models complex multivariate extremes using geometric shapes.

Extends geometric approach to model non-stationary extremal dependence.

This paper uses VAE to generate extreme events from multivariate data.

The paper introduces twins in Kähler and Sasaki geometry, generalizing known concepts.

We study the Sasaki cone of a CR structure of Sasaki type on a given closed manifold. We introduce an energy functional over the cone, and use its critical points to single out the strongly extremal Reeb vectors fields. Should one such vector field be a member of the extremal set, the scalar curvature of a Sasaki extre…

New bandit algorithms focus on extreme values, outperforming existing methods.

Algorithms for hyperparameter optimization abound, all of which work well under different and often unverifiable assumptions. Motivated by the general challenge of sequentially choosing which algorithm to use, we study the more specific task of choosing among distributions to use for random hyperparameter optimization.…

We apply the theory of continuous time random walks to study some aspects of the extreme value problem applied to financial time series. We focus our attention on extreme times, specifically the mean exit time and the mean first-passage time. We set the general equations for these extremes and evaluate the mean exit ti…

Study on extremal subsets in geodesically complete spaces with curvature constraints.

This paper is a survey of some recent progress on the study of Calabi's extremal Kähler metrics. We first discuss the Yau-Tian-Donaldson conjecture relating the existence of extremal metrics to an algebro-geometric stability notion and we give some example settings where this conjecture has been established. We then tu…

We define regular points of an extremal subset in an Alexandrov space and study their basic properties. We show that a neighborhood of a regular point in an extremal subset is almost isometric to an open subset in Euclidean space and that the set of regular points in an extremal subset has full measure and is dense in …

Proves minimization for Kähler manifolds with automorphisms.

This paper uses MIS to identify key financial institutions with minimal risk contagion.

Survey of extreme value modeling techniques for insurance.

Improved Frank-Wolfe algorithm for constrained convex optimization with nearest extreme point oracle.

Paper finds robust $Λ$-quantiles equal to extremal distributions.

Study compares and accelerates deep learning for extreme events modeling.

Extends fractional $L^p$ uncertainty principles with extremizers and stability results.

New method estimates root-directed tree from extreme data.

Proposes a network-based strategy to manage financial market risks.

We investigate the supports of extremal martingale measures with pre-specified marginals in a two-period setting. First, we establish in full generality the equivalence between the extremality of a given measure $Q$ and the denseness in $L^1(Q)$ of a suitable linear subspace, which can be seen in a financial context as…

Classification tasks usually assume that all possible classes are present during the training phase. This is restrictive if the algorithm is used over a long time and possibly encounters samples from unknown classes. The recently introduced extreme value machine, a classifier motivated by extreme value theory, addresse…

New method models precipitation extremes and spatial dependence.

The study introduces new liquidity measures and models for assets with extreme liquidity.

Proposes a method to model financial returns with extreme shocks using flexible tail transformations.

Defines new extremal potentials and measures for Kähler forms.

We present a novel distribution-free approach, the data-driven threshold machine (DTM), for a fundamental problem at the core of many learning tasks: choose a threshold for a given pre-specified level that bounds the tail probability of the maximum of a (possibly dependent but stationary) random sequence. We do not ass…

Extreme learning machine (ELM) is a new single hidden layer feedback neural network. The weights of the input layer and the biases of neurons in hidden layer are randomly generated, the weights of the output layer can be analytically determined. ELM has been achieved good results for a large number of classification ta…

Recently it was shown that the area A and the angular momentum J of any apparent horizon on a maximal, axisymmetric and asymptotically flat Cauchy hyper-surface of a vacuum space-time satisfy necessarily the universal inequality A >= 8 pi |J|. We show here that the equality A=8 pi |J| is never attained. As equality is …

New method uses neural networks to predict extreme wildfires, improving accuracy over traditional models.

Combines GANs and EVT for better modeling of spatial climate extremes.

Based on recent work of S. K. Donaldson and T. Mabuchi, we prove that any extremal Kaehler metric in the sense of E. Calabi, defined on the product of polarized compact complex projective manifolds is the product of extremal Kaehler metrics on each factor, provided that the integral Futaki invariants of the polarized m…

Study examines dependence of extreme electricity prices in Australian markets.

Taxicab correspondence analysis visualizes sparse text data sets.

Quantile regression is an increasingly important empirical tool in economics and other sciences for analyzing the impact of a set of regressors on the conditional distribution of an outcome. Extremal quantile regression, or quantile regression applied to the tails, is of interest in many economic and financial applicat…

Alexandrov spaces have a special stratification that maps to spheres.

A new algorithm for selecting top-k arms in extreme contextual bandits with improved efficiency.

Extreme value theory enhances statistical learning extrapolation for rare events.

Study on Heisenberg group's Lorentzian problems using Pontryagin's principle.

Inference over tails is usually performed by fitting an appropriate limiting distribution over observations that exceed a fixed threshold. However, the choice of such threshold is critical and can affect the inferential results. Extreme value mixture models have been defined to estimate the threshold using the full dat…

New method simulates multivariate extreme events using GANs and Aitchison coordinates.

We introduce the notion of partial presimplicial set and construct its geometric realization. We show that any semiadequate diagram yields a partial presimplicial set leading to a geometric realization of the almost-extreme Khovanov homology of the diagram. We give a concrete formula for the homotopy type of this geome…

Researchers solved a problem about extreme mass distributions in quasi-copulas.

We generalize to the finite-state case the notion of the extreme effect variable $Y$ that accumulates all the effect of a variant variable $V$ observed in changes of another variable $X$. We conduct theoretical analysis and turn the problem of finding of an effect variable into a problem of a simultaneous decomposition…

Extreme multi-label text classification (XMTC) addresses the problem of tagging each text with the most relevant labels from an extreme-scale label set. Traditional methods use bag-of-words (BOW) representations without context information as their features. The state-ot-the-art deep learning-based method, AttentionXML…

New findings extend rigidity results to broader classes of manifolds.

The novel unseen classes can be formulated as the extreme values of known classes. This inspired the recent works on open-set recognition \cite{Scheirer_2013_TPAMI,Scheirer_2014_TPAMIb,EVM}, which however can have no way of naming the novel unseen classes. To solve this problem, we propose the Extreme Value Learning (E…

We introduce a family of extremal polynomials associated with the prolongation of a stratified nilpotent Lie algebra. These polynomials are related to a new algebraic characterization of abnormal subriemannian geodesics in stratified nilpotent Lie groups. They satisfy a set of remarkable structure relations that are us…