Density-based outliers in labor market text act as leading indicators of new occupational clusters, with an extended Emerging Occupation Score predicting formation 2 quarters ahead at F1=0.74 on 84,988 postings.
Breunig, Hans-Peter Kriegel, Raymond T
9 Pith papers cite this work. Polarity classification is still indexing.
years
2026 9representative citing papers
Latent SDE generative model for anomaly detection in sparse irregular multivariate time series outperforms baselines on six benchmarks and stays robust under severe sparsity.
A metric learning method is introduced to learn distance metrics that best capture conditional anomaly patterns in instance-based detection.
An LLM agent automates iterative refinement of data embedding visualizations by generating semantic evaluation reports and recommending configuration changes.
Four-planet systems exhibit exponentially increasing lifetimes with orbital spacing, intermediate between three- and five-planet systems, with resonances causing shorter lifetimes and third-order MMRs adding destabilization near certain spacings.
Unsupervised UMAP plus Friends-of-Friends clustering on 58 million DESI DR2 spectra identifies 1.1 million outlier candidates where visual checks show 67 percent have reduction artifacts missed by the standard pipeline.
Ensemble anomaly detection framework for real-time risk calculation monitoring outperforms single methods with F1 scores of 61-79% on proprietary credit-derivatives data using injected anomalies.
Instance-based conditional anomaly detection with optimized distance metrics detects unusual patient-management decisions in two real-world medical datasets.
GNN embeddings turn CEBAF injector snapshots into a coordinate system revealing ten persistent operating regimes that support monitoring, outlier detection, and case-based reasoning over a year of data.
citing papers explorer
-
Noise is Signal: Density-Based Outliers as Leading Indicators of Occupational Emergence in Labor Market Text
Density-based outliers in labor market text act as leading indicators of new occupational clusters, with an extended Emerging Occupation Score predicting formation 2 quarters ahead at F1=0.74 on 84,988 postings.
-
Anomaly Detection for Sparse and Irregular Multivariate Time Series with Latent SDEs
Latent SDE generative model for anomaly detection in sparse irregular multivariate time series outperforms baselines on six benchmarks and stays robust under severe sparsity.
-
Distance metric learning for conditional anomaly detection
A metric learning method is introduced to learn distance metrics that best capture conditional anomaly patterns in instance-based detection.
-
Explainable Iterative Data Visualisation Refinement via an LLM Agent
An LLM agent automates iterative refinement of data embedding visualizations by generating semantic evaluation reports and recommending configuration changes.
-
Orbital Stability of Closely-Spaced Four-planet Systems
Four-planet systems exhibit exponentially increasing lifetimes with orbital spacing, intermediate between three- and five-planet systems, with resonances causing shorter lifetimes and third-order MMRs adding destabilization near certain spacings.
-
Quality Assessment of Spectroscopic Data Reduction Pipelines Using Artificial Intelligence: Scrutinizing Data Release 2 from the DESI Survey
Unsupervised UMAP plus Friends-of-Friends clustering on 58 million DESI DR2 spectra identifies 1.1 million outlier candidates where visual checks show 67 percent have reduction artifacts missed by the standard pipeline.
-
How to spot outliers: an Ensemble Anomaly Detection Framework
Ensemble anomaly detection framework for real-time risk calculation monitoring outperforms single methods with F1 scores of 61-79% on proprietary credit-derivatives data using injected anomalies.
-
Conditional anomaly detection methods for patient-management alert systems
Instance-based conditional anomaly detection with optimized distance metrics detects unusual patient-management decisions in two real-world medical datasets.
-
Machine-State Embeddings as an Operational Coordinate System for Accelerator Operation
GNN embeddings turn CEBAF injector snapshots into a coordinate system revealing ten persistent operating regimes that support monitoring, outlier detection, and case-based reasoning over a year of data.