Time Series Data Augmentation for Deep Learning: A Survey

Abstract

Deep learning performs remarkably well on many time series analysis tasks recently. The superior performance of deep neural networks relies heavily on a large number of training data to avoid overfitting. However, the labeled data of many real-world time series applications may be limited such as classification in medical time series and anomaly detection in AIOps. As an effective way to enhance the size and quality of the training data, data augmentation is crucial to the successful application of deep learning models on time series data. In this paper, we systematically review different data augmentation methods for time series. We propose a taxonomy for the reviewed methods, and then provide a structured review for these methods by highlighting their strengths and limitations. We also empirically compare different data augmentation methods for different tasks including time series classification, anomaly detection, and forecasting. Finally, we discuss and highlight five future directions to provide useful research guidance.

Accepted by the 30th International Joint Conference on Artificial Intelligence (IJCAI 2021); Selected by Paper Digest into Most Influential IJCAI Papers (Version: 2022-02), Rank 1st (Link: https://www.paperdigest.org/2022/02/most-influential-ijcai-papers-2022-02/)

Cited by in corpus (22)

• An Empirical Survey of Data Augmentation for Time Series Classification with Neural Networks
• A Survey on Data Augmentation for Text Classification
• Data Augmentation techniques in time series domain: A survey and taxonomy
• Self-supervised Contrastive Representation Learning for Semi-supervised Time-Series Classification
• TFAD: A Decomposition Time Series Anomaly Detection Architecture with Time-Frequency Analysis
• COCOA: Cross Modality Contrastive Learning for Sensor Data
• Reliability and Interpretability in Science and Deep Learning
• Label-efficient Time Series Representation Learning: A Review
• DACAD: Domain Adaptation Contrastive Learning for Anomaly Detection in Multivariate Time Series
• Automated scoring of pre-REM sleep in mice with deep learning
• Time Series Forecasting via Learning Convolutionally Low-Rank Models
• Fractal interpolation in the context of prediction accuracy optimization
• Deep Learning Methods for Abstract Visual Reasoning: A Survey on Raven's Progressive Matrices
• Data Augmentation Methods of Dynamic Model Identification for Harbor Maneuvers using Feedforward Neural Network
• Topological Persistence Guided Knowledge Distillation for Wearable Sensor Data
• Balance Measures Derived from Insole Sensor Differentiate Prodromal Dementia with Lewy Bodies
• RHiOTS: A Framework for Evaluating Hierarchical Time Series Forecasting Algorithms
• Deep Convolutional Autoencoder for Assessment of Drive-Cycle Anomalies in Connected Vehicle Sensor Data
• A comparative study of non-deep learning, deep learning, and ensemble learning methods for sunspot number prediction
• DTWSSE: Data Augmentation with a Siamese Encoder for Time Series
• An Empirical Evaluation of Multivariate Time Series Classification with Input Transformation across Different Dimensions
• PIE-ADA: Physics-Informed Ensemble with Adaptive Data Augmentation for Photometric Transient Classification