anomaly-detection

This paper analyzes failures in reconstruction-based unsupervised learning through geometric conditions and proposes new methods like Dynamic Push and Pull to enhance anomaly detection performance.

This paper introduces a framework using AUC bounds as a differentiable objective for anomaly-free self-optimization of anomaly detection systems, achieving performance gains over conventional model selection methods.

This paper presents Signal2Symbol, a neuro-symbolic framework for explainable anomaly detection in physiological time-series data like ECG and EEG, leveraging symbolic tokenization and temporal reasoning to produce interpretable explanations of anomalous patterns.

The paper proposes UGCAD, an unsupervised framework using variational graph autoencoders and clustering to detect collective anomalies in IoT network traffic, with experiments on benchmark datasets showing its effectiveness.

DiFA introduces a dual-evidence fusion and aggregation framework for token-level text anomaly detection, improving performance by combining form-structural and semantic views across various benchmarks.

World Signal is a product that monitors real-world systems like earthquakes, wildfires, internet connectivity, and aviation to detect unusual activity across the planet, allowing users to explore anomalies and compare them with historical data.

Agent Anomaly Detection is a new feature in Private Preview on the Gemini Enterprise Agent Platform that provides reasoning-based oversight for AI agents, detecting anomalies and policy violations asynchronously without runtime latency.

Anomalo Analyst proactively monitors data in Snowflake, Databricks, or BigQuery, surfaces important trends and anomalies, and allows investigation with plain language queries, with insights verified against the data.

This paper proposes a deep reinforcement learning framework for risk and anomaly identification in distribution networks, using uncertainty quantification to distinguish between inherent risks and out-of-distribution anomalies.

RankShift is a novel in-database method for detecting and explaining categorical shifts in data streams, using Pearson scores to identify responsible categories, and it shows competitive performance against autoencoders on log datasets.

A survey paper that categorizes LLM-based agents for time-series tasks into four categories and provides a task-oriented guide for design and future research.

TRACE-C is a rank-calibrated detector for identifying joint anomalies in multi-stream operational telemetry, evaluated on power grid data.

The paper proposes OTTA-DGAD, a method for online test-time adaptation in dynamic graph anomaly detection that uses dynamic prototypes and memory buffers to handle unseen target domains without retraining.

This paper investigates unsupervised anomaly detection using flow matching on tabular data, focusing on contaminated training sets and comparing different scoring methods for robustness.

The paper proposes LLM-Detector, a framework that uses large language models with in-context learning to perform tabular anomaly detection without fine-tuning, demonstrating consistent improvements over existing methods on multiple datasets.

The paper proposes CMR-Mamba, a causal mechanism monitoring method using Mamba state-space models for unsupervised industrial fault detection, focusing on coupling faults and evaluated on electromechanical, hydraulic, and cyber-physical systems.

This paper revisits energy-based models for tabular anomaly detection, demonstrating that combining Deep Boltzmann Machine energy scores with autoencoder reconstruction scores significantly improves performance on benchmark datasets.

ORCA is a two-stage framework using contrastive learning and autoencoders for interpretable anomaly detection in collider experiments, enhancing sensitivity and interpretability for new physics searches.

This paper introduces X-AddGraph, a post-hoc explainability framework for the AddGraph dynamic graph anomaly detector, using a dual spatial-temporal attribution mechanism that preserves detection performance exactly while providing explanations.

This paper evaluates different multivariate outlier detection methods (Z-score, Mahalanobis distances, PCA, Isolation Forest, and Hotelling's T-squared) for identifying irregular operations in district heating system data, proposing an ensemble approach based on agreement among the best-performing methods.