ExGAN generates realistic extreme samples using GANs and EVT.
problem Generating realistic extreme scenarios for risk management.
method ExGAN combines GANs with EVT to model extreme tails of distributions.
result ExGAN efficiently generates extreme samples with constant time complexity.
Spectral clustering identifies clusters of multivariate extremes.
problem Analyzing the dependence structure of multivariate extremes.
method Spectral clustering based on a random k-nearest neighbor graph. result Spectral clustering can consistently identify clusters of multivariate extremes under certain conditions.
This paper uses VAE to generate extreme events from multivariate data.
problem Generating accurate extremes from observational data for risk assessment.
method Variational Autoencoder (VAE) approach for multivariate heavy-tailed distributions.
result Improves learning of dependency structure between extremes.
Two BO methods improve reliability optimization for rare failures.
problem Maximizing reliability of designs subject to random perturbations.
method Bayesian optimization with Thompson sampling and knowledge gradient.
result Proposed methods outperform existing techniques in extreme failure probability scenarios.
Paper introduces SPADE method to protect classifiers from OOD and adversarial samples.
problem Protecting classifiers from out-of-distribution and adversarial samples.
method SPADE method based on GEV model in latent space.
result Provable protection against OOD and adversarial samples.
Being able to predict the occurrence of extreme returns is important in financial risk management. Using the distribution of recurrence intervals---the waiting time between consecutive extremes---we show that these extreme returns are predictable on the short term. Examining a range of different types of returns and th…
The paper examines extreme value statistics of high-dimensional sample covariances, with applications in finance and image analysis.
problem Statistical validation of normal conditions in high-dimensional time series data.
method Generalizes the maximal deviation of sample autocovariances to high dimensions and applies Gumbel-type extreme value asymptotics.
result Gumbel-type extreme value asymptotics holds true for high-dimensional sample covariances.
Kernel PCA helps analyze multivariate extremes and clusters them effectively.
problem Analyzing the dependence structure of multivariate extremes.
method Kernel PCA as a method for clustering and dimension reduction.
result Kernel PCA preimages effectively identify clusters in multivariate extremes.
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…
GARCH-UGH improves VaR estimation for financial risk management.
problem Dynamic estimation of extreme VaR in financial time series.
method AR-GARCH filtering followed by a bias-reduced extreme value estimator.
result GARCH-UGH estimates are more accurate than conventional methods.
We assess cluster stability by trimming extreme points and tracking data range reduction.
problem Assessing stability of one-dimensional clusters.
method Probabilistic method using diameter-shrinkage ratio to track data range reduction.
result Our method achieves higher accuracy than classical tests in small or noisy samples.
New model predicts financial tail events using RIA-EVT-Copula.
problem Predicting financial tail events for risk management.
method RIA-EVT-Copula framework combining POT, RIA, and copulas.
result Improved accuracy in predicting financial extremes.
A new method reduces uncertainty in predicting rare extreme events without assuming their presence in training data.
problem Predicting rare and extreme events in complex systems with high uncertainty.
method Extreme Event Aware (e2a or η) learning, which enforces extreme event statistics during training.
result Models generate unprecedented extreme events even when training data lacks extremes.
Improved analysis for extreme multi-class CRL with better sample complexity.
problem Theoretical sample complexity of CRL in extreme multi-class settings is poorly understood.
method Improved U-Statistics estimator to capture class concentration, proving O(k) sample complexity. result Sample complexity is O(k) for extreme multi-class learning, independent of class distribution. We give some detailed numerical information about extremal metrics on four different toric surfaces. These are sample of many other cases which can be treated using a computer programme outlined in the paper.
This paper tackles label-efficient evaluation in extreme class imbalance.
problem Challenges in obtaining a sufficient sample for accurate evaluation in tasks with extreme class imbalance.
method Develops a framework for online evaluation based on adaptive importance sampling.
result Establishes strong consistency and a central limit theorem for performance estimates.
The paper proposes a new variant of a decision tree, called an Extreme Learning Tree. It consists of an extremely random tree with non-linear data transformation, and a linear observer that provides predictions based on the leaf index where the data samples fall. The proposed method outperforms linear models on a bench…
New bandit algorithms focus on extreme values, outperforming existing methods.
problem Optimizing decisions based on extreme values rather than expected values.
method Robust statistics-based algorithms with vanishing extremal regret.
result The proposed algorithms achieve superior performance compared to existing methods.
Deep learning models complex multivariate extremes using geometric shapes.
problem Modeling complex extremal dependencies in high-dimensional data.
method Geometric representation and deep learning for flexible semi-parametric models.
result First approach to modeling limit sets using deep learning for high-dimensional data.
New method reduces bias in learning from large action spaces using selective importance sampling.
problem Learning from large-scale recommendation systems with bandit feedback and supervised labels.
method Selective Importance Sampling (sIS) and Policy Optimization for eXtreme Models (POXM) algorithm.
result POXM method significantly outperforms existing methods in learning from bandit feedback on XMC tasks.
AI boosts study of rare weather extremes with lower costs.
problem Difficulty in studying rare weather events due to limited data and models.
method Coupling AI forecasts with physics models using rare-event algorithms.
result Efficiently characterizes very rare events like once-per-millennium heatwaves.
Combines GANs and EVT for better modeling of spatial climate extremes.
problem Modeling dependencies between climate extremes, especially in high-dimensional spaces.
method Generative Adversarial Networks (GANs) combined with Extreme Value Theory (EVT).
result evtGAN outperforms classical GANs and statistical approaches in modeling spatial extremes.
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…
The paper provides bounds for the empirical angular measure and applies them to improve statistical learning in extreme regions.
problem Estimating the angular measure in high-dimensional data with different distributions.
method Established bounds for the maximal deviations of the empirical angular measure from the true measure, using rank transformation and analyzing the most extreme observations.
result The bounds provide performance guarantees for statistical learning procedures in extreme regions, such as binary classification and anomaly detection.
CEDA analyzes large categorical datasets using tree geometry and binary codes.
problem Analyzing large categorical datasets with extreme-K samples. method CEDA uses tree geometry and binary codes to analyze categorical data.
result CEDA discovers patterns and evaluates their reliability in large categorical datasets.
Improved Thompson Sampling for Bayesian Optimization.
problem Handling the exploitation-exploration dilemma in Bayesian optimization.
method Incorporating epsilon-greedy policy into Thompson Sampling.
result Epsilon-greedy Thompson Sampling outperforms standard TS extremes.
Framework reconstructs missing spatio-temporal data for extreme value prediction.
problem Predicting extreme values from incomplete spatio-temporal data.
method Convolutional deep neural networks and autoencoder-like models for conditional sampling.
result Framework produces accurate reconstructions of missing data for extremal values.
Study optimizes sampling to avoid extreme tail risks in unknown heavy-tailed distributions.
problem Identify optimal alternative with minimal extreme tail risk from unknown heavy-tailed distributions.
method Data-driven sequential sampling policies to maximize likelihood of selecting the optimal alternative.
result Proposed methods outperform existing approaches in identifying the optimal alternative.
Extends geometric approach to model non-stationary extremal dependence.
problem Capturing evolving extremal dependence in multivariate data.
method Geometric framework for non-stationary multivariate extreme value modelling.
result Framework can capture various dependence forms and is robust to different model formulations.
Prediction intervals in supervised Machine Learning bound the region where the true outputs of new samples may fall. They are necessary in the task of separating reliable predictions of a trained model from near random guesses, minimizing the rate of False Positives, and other problem-specific tasks in applied Machine …
The hidden tail of empirical distributions is analyzed using extreme value theory.
problem Understanding the bias between in-sample mean and true statistical mean for large n. method Extreme value theory applied to empirical distributions and their moments.
result The hidden moment of order 0 for power law distributions follows an exponential distribution with expectation 1/n. This paper addresses the problem of estimating, in the presence of random censoring as well as competing risks, the extreme value index of the (sub)-distribution function associated to one particular cause, in the heavy-tail case. Asymptotic normality of the proposed estimator (which has the form of an Aalen-Johansen i…
Develops statistical framework for analyzing functional data extremes.
problem Analyzing extremes of functional data in Hilbert spaces.
method Regular variation in Hilbert spaces, Peaks-Over-Threshold framework, functional PCA.
result Proposes a dimension reduction method for functional extreme observations.
New method models precipitation extremes and spatial dependence.
problem Estimating dependencies of precipitation maxima in space and time.
method Generative neural networks for max-stable processes.
result Explicit nonparametric estimate of spatial dependence.
New method simulates multivariate extreme events using GANs and Aitchison coordinates.
problem Simulating multivariate extreme events for economic risk assessment.
method Wasserstein-Aitchison GAN approach combining tail dependence and marginal tail modeling.
result Strong performance in capturing tail dependence and generating accurate extreme observations.
Paper proposes efficient GCN learning method for limited data.
problem Learning GCNs from data with extremely limited annotations.
method Adaptive sampling strategy and model compression.
result Cut down annotation requirement by 90% and compress parameters 6x.
In a wide variety of sequential decision making problems, it can be important to estimate the impact of rare events in order to minimize risk exposure. A popular risk measure is the conditional value-at-risk (CVaR), which is commonly estimated by averaging observations that occur beyond a quantile at a given confidence…
Study on consistency of ML methods for moving objects in non-stationary environments.
problem Consistency of machine learning methods for moving objects in non-stationary environments.
method Least squares, ridge regression, and ℓs-penalized least squares methods under non-stationary spatial-temporal sampling. result Consistency and asymptotic normality of the estimates under weak conditions.
New acquisition function for extreme rewards in bandits.
problem Online decision making with extreme payoffs in multi-armed bandits.
method Modeling payoffs as Gaussian processes and using a novel UCB acquisition function.
result Demonstrated benefits across synthetic and real-world benchmarks.
New method learns graphical models with latent variables for extreme events.
problem Learning graphical models with latent variables for multivariate extremes.
method Tractable convex program exttt{eglatent} for Hüsler-Reiss models.
result Consistently recovers conditional graph and latent variables.
We develop a method for the evaluation of extreme event statistics associated with nonlinear dynamical systems, using a small number of samples. From an initial dataset of design points, we formulate a sequential strategy that provides the 'next-best' data point (set of parameters) that when evaluated results in improv…
Novel SVM approach for extreme quantile regression with heavy tailed inputs.
problem Learning from extreme values in quantile regression.
method Support Vector Machine framework for handling high-dimensional and nonlinear settings.
result Established finite-sample learning guarantees under mild regularity assumptions.
Training a classifier over a large number of classes, known as 'extreme classification', has become a topic of major interest with applications in technology, science, and e-commerce. Traditional softmax regression induces a gradient cost proportional to the number of classes C, which often is prohibitively expensive…
The Extreme Deconvolution method fits a probability density to a dataset where each observation has Gaussian noise added with a known sample-specific covariance, originally intended for use with astronomical datasets. The existing fitting method is batch EM, which would not normally be applied to large datasets such as…
Improved Hawkes model forecasts extreme financial returns more accurately.
problem Forecasting extreme tail events in financial log-returns.
method 2T-POT Hawkes model with multiple exceedance thresholds.
result 2T-POT Hawkes model outperforms GARCH-EVT model in risk forecasting.
Study tail risk in high-frequency finance using L1-regularized regression.
problem Measuring tail risk dynamics in high-frequency financial markets.
method Dynamic extreme value regression model with L1-regularized maximum likelihood estimator. result Severity of extreme losses well predicted by low price impact in high volatility periods.
This paper develops DRO estimators for EVT statistics using point processes.
problem Scarcity of extreme data leads to model misspecification error in EVT.
method Developed DRO estimators informed by semi-parametric max-stable constraints in the space of point processes.
result Proposed DRO estimators improve out-of-sample performance and are validated on synthetic and real data.
By decomposing asset returns into potential maximum gain (PMG) and potential maximum loss (PML) with price extremes, this study empirically investigated the relationships between PMG and PML. We found significant asymmetry between PMG and PML. PML significantly contributed to forecasting PMG but not vice versa. We furt…