Precision and Recall for Time Series

TL;DR

This paper introduces a new mathematical model that extends Precision and Recall metrics to evaluate range-based anomalies in time series, accommodating domain-specific customization.

Contribution

It presents a novel model for assessing time series classification accuracy by adapting Precision and Recall for range-based anomalies with customizable options.

Findings

Model effectively measures range-based anomaly detection accuracy

Supports domain-specific customization of evaluation metrics

Enhances traditional point-based metrics for time series analysis

Abstract

Classical anomaly detection is principally concerned with point-based anomalies, those anomalies that occur at a single point in time. Yet, many real-world anomalies are range-based, meaning they occur over a period of time. Motivated by this observation, we present a new mathematical model to evaluate the accuracy of time series classification algorithms. Our model expands the well-known Precision and Recall metrics to measure ranges, while simultaneously enabling customization support for domain-specific preferences.