Conditional anomaly detection using soft harmonic functions: An application to clinical alerting

2026-04-23

AI Summary

This work addresses the challenge of label misannotation in clinical practice, where anomalous events often manifest as omissions of critical procedures such as laboratory tests. The authors propose a nonparametric conditional anomaly detection method based on a soft harmonic function to estimate label confidence and identify mislabeled instances. To enhance robustness, the approach incorporates distributional support boundary regularization, which prevents isolated or marginally located samples from being erroneously classified as anomalies. Requiring no strong parametric assumptions about the underlying model, the method demonstrates significant performance gains over existing baselines on real-world electronic health record data, thereby improving the reliability and practical utility of anomaly detection in clinical settings.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

Timely detection of concerning events is an important problem in clinical practice. In this paper, we consider the problem of conditional anomaly detection that aims to identify data instances with an unusual response, such as the omission of an important lab test. We develop a new non-parametric approach for conditional anomaly detection based on the soft harmonic solution, with which we estimate the confidence of the label to detect anomalous mislabeling. We further regularize the solution to avoid the detection of isolated examples and examples on the boundary of the distribution support. We demonstrate the efficacy of the proposed method in detecting unusual labels on a real-world electronic health record dataset and compare it to several baseline approaches.

Problem

• conditional anomaly detection
• clinical alerting
• unusual response
• anomalous mislabeling
• electronic health record

Innovation

• conditional anomaly detection
• soft harmonic functions
• non-parametric method
• label confidence
• clinical alerting