Overview

Extreme Value Theory studies the behavior of unusually large or small observations, helping estimate rare-event probabilities and extreme quantiles when data are scarce. Common approaches model threshold exceedances or block maxima, estimate tail shape, and account for dependence; newer work combines these tools with boosting, kernel methods, generative models, and causal inference. Applications include insurance losses, communications, spatial systems, and safety-critical prediction. A central challenge is that tail estimates can be sensitive to modeling assumptions and threshold choices, so reliable extrapolation needs diagnostics and should sometimes be withheld when the data provide too little support.

Latest papers

CardsList

15 papers3 papers in the last four weeks, with none the four weeks before. 0.0% of all new papers.

  1. Generative Modeling for Cascading Extremes via Autoregressive Transformers

    Nov 3, 2026Miguel de Carvalho, Myung Won Lee, Jordan Richards +1Generative ModelingTime Series Tokenization

  2. FADE: A Flow-Based Approach for Multichannel Tail Estimation

    Nov 1, 2026Parmida Valiahdi, Niloofar Mehrnia, Mustafa Furkan Beker +3Rare-Event EstimationExtreme Value Theory

  3. Deep catastrophic-risk-aware reinforcement learning

    Oct 19, 2026Carlos Octavio Pérez Mendoza, Frédéric GodinRisk-Sensitive RLConditional Value-at-Risk

  4. A Location-Invariant Estimator of Extremal Quantile Treatment Effects for Heavy-Tailed Distributions

    Sep 3, 2026Xin Yu, Shuwei Huang, Jicheng Liu +4Causal Effect EstimationHeavy-Tailed Distributions

  5. Physics-informed VAE-EVT for Tail Aware Radio Map Prediction

    Aug 15, 2026Amanda Sheron Gamage, Niloofar Mehrnia, James GrossRadio Propagation ModelingWireless Communications

  6. Gradient boosting for extremes: sampling theory and application to insurance

    Jun 12, 2026Stéphane Lhaut, Olivier LopezGradient-Boosted Decision TreesHeavy-Tailed Distributions

  7. Out-of-Distribution generalization of quantile regression with heavy tailed inputs: an SVM approach

    May 29, 2026Baptiste Leroux, Clément Dombry, Anne SabourinQuantile RegressionOOD Generalization

  8. When Softmax Fails at the Top: Extreme Value Corrections for InfoNCE

    May 29, 2026Melihcan Erol, Suat Evren, Oktay Ozel +3Contrastive LearningRepresentation Learning

  9. Extrapolation in Statistical Learning with Extreme Value Theory

    May 3, 2026Sebastian Engelke, Nicola Gnecco, Anne SabourinOOD GeneralizationStatistical Learning Theory

  10. EVT-Based Generative AI for Tail-Aware Channel Estimation

    Apr 27, 2026Parmida Valiahdi, Niloofar Mehrnia, Walid Saad +1Synthetic Data AugmentationWireless Communications

  11. Multi-output Extreme Spatial Model for Complex Aircraft Production Systems

    Apr 24, 2026Cheolhei Lee, Xing Wang, Xiaowei Yue +1Extreme Value Theory

  12. High-Dimensional Tail Index Regression

    Mar 2, 2024Yuya Sasaki, Jing Tao, Yulong WangHeavy-Tailed DistributionsExtreme Value Theory