Hybrid Efficient Unsupervised Anomaly Detection for Early Pandemic Case Identification

TL;DR

This paper presents a hybrid unsupervised anomaly detection method combining distance and density measures, demonstrating improved early pandemic case identification using COVID-19 chest X-ray data.

Contribution

It introduces a novel hybrid anomaly detection approach that outperforms existing methods in early epidemic case detection, especially with limited data.

Findings

Achieved an average AUC of 77.43% on COVID-19 X-ray data

Outperformed Isolation Forest and KNN in anomaly detection accuracy

Proven effective in early pandemic scenario detection

Abstract

Unsupervised anomaly detection is a promising technique for identifying unusual patterns in data without the need for labeled training examples. This approach is particularly valuable for early case detection in epidemic management, especially when early-stage data are scarce. This research introduces a novel hybrid method for anomaly detection that combines distance and density measures, enhancing its applicability across various infectious diseases. Our method is especially relevant in pandemic situations, as demonstrated during the COVID-19 crisis, where traditional supervised classification methods fall short due to limited data. The efficacy of our method is evaluated using COVID-19 chest X-ray data, where it significantly outperforms established unsupervised techniques. It achieves an average AUC of 77.43%, surpassing the AUC of Isolation Forest at 73.66% and KNN at 52.93%. These results highlight the potential of our hybrid anomaly detection method to improve early detection capabilities in diverse epidemic scenarios, thereby facilitating more effective and timely responses.