Conditional anomaly detection with soft harmonic functions

Abstract

In this paper, we consider the problem of conditional anomaly detection that aims to identify data instances with an unusual response or a class label. 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 on several synthetic and UCI ML datasets in detecting unusual labels when compared to several baseline approaches. We also evaluate the performance of our method on a real-world electronic health record dataset where we seek to identify unusual patient-management decisions.

Published at IEEE International Conference on Data Mining (ICDM 2011). 10.1109/ICDM.2011.40

References in corpus (7)

• Identifying Mislabeled Training Data
• Estimating time-varying networks
• Bayesian anomaly detection methods for social networks
• Conditional outlier detection for clinical alerting
• Evidence-based anomaly detection in clinical domains
• Conditional anomaly detection methods for patient-management alert systems
• Distance metric learning for conditional anomaly detection

Cited by in corpus (7)

• Artificial Intelligence based Anomaly Detection of Energy Consumption in Buildings: A Review, Current Trends and New Perspectives
• On the Nature and Types of Anomalies: A Review of Deviations in Data
• Explainable Contextual Anomaly Detection using Quantile Regression Forests
• Detecting Outliers in Data with Correlated Measures
• Robust Contextual Outlier Detection: Where Context Meets Sparsity
• MCODE: Multivariate Conditional Outlier Detection
• Robust Subspace Outlier Detection in High Dimensional Space