Conditional anomaly detection with soft harmonic functions
Valko M, Kveton B, Valizadegan H, Cooper GF, Hauskrecht M. Conditional anomaly detection with soft harmonic functions. In: Proceedings of the International Conference on Data Mining (2011) 735-743.
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.