An Automated Machine Learning Approach for Detecting Anomalous Peak Patterns in Time Series Data from a Research Watershed in the Northeastern United States Critical Zone

TL;DR

This paper introduces an automated machine learning framework that detects peak-pattern anomalies in watershed sensor data by generating synthetic datasets, optimizing model selection among five deep learning architectures based on accuracy and computational cost.

Contribution

The framework automates anomaly detection model selection in time series data by integrating synthetic data generation, hyperparameter optimization, and multi-model evaluation tailored to user preferences.

Findings

Framework effectively identifies optimal models for anomaly detection.

Synthetic data generation improves model training and evaluation.

Model selection aligns with user-defined accuracy and computational constraints.

Abstract

This paper presents an automated machine learning framework designed to assist hydrologists in detecting anomalies in time series data generated by sensors in a research watershed in the northeastern United States critical zone. The framework specifically focuses on identifying peak-pattern anomalies, which may arise from sensor malfunctions or natural phenomena. However, the use of classification methods for anomaly detection poses challenges, such as the requirement for labeled data as ground truth and the selection of the most suitable deep learning model for the given task and dataset. To address these challenges, our framework generates labeled datasets by injecting synthetic peak patterns into synthetically generated time series data and incorporates an automated hyperparameter optimization mechanism. This mechanism generates an optimized model instance with the best architectural and training parameters from a pool of five selected models, namely Temporal Convolutional Network (TCN), InceptionTime, MiniRocket, Residual Networks (ResNet), and Long Short-Term Memory (LSTM). The selection is based on the user's preferences regarding anomaly detection accuracy and computational cost. The framework employs Time-series Generative Adversarial Networks (TimeGAN) as the synthetic dataset generator. The generated model instances are evaluated using a combination of accuracy and computational cost metrics, including training time and memory, during the anomaly detection process. Performance evaluation of the framework was conducted using a dataset from a watershed, demonstrating consistent selection of the most fitting model instance that satisfies the user's preferences.