Deep learning for time series classification: a review

Abstract

Time Series Classification (TSC) is an important and challenging problem in data mining. With the increase of time series data availability, hundreds of TSC algorithms have been proposed. Among these methods, only a few have considered Deep Neural Networks (DNNs) to perform this task. This is surprising as deep learning has seen very successful applications in the last years. DNNs have indeed revolutionized the field of computer vision especially with the advent of novel deeper architectures such as Residual and Convolutional Neural Networks. Apart from images, sequential data such as text and audio can also be processed with DNNs to reach state-of-the-art performance for document classification and speech recognition. In this article, we study the current state-of-the-art performance of deep learning algorithms for TSC by presenting an empirical study of the most recent DNN architectures for TSC. We give an overview of the most successful deep learning applications in various time series domains under a unified taxonomy of DNNs for TSC. We also provide an open source deep learning framework to the TSC community where we implemented each of the compared approaches and evaluated them on a univariate TSC benchmark (the UCR/UEA archive) and 12 multivariate time series datasets. By training 8,730 deep learning models on 97 time series datasets, we propose the most exhaustive study of DNNs for TSC to date.

Accepted at Data Mining and Knowledge Discovery

References in corpus (32)

• Efficient Estimation of Word Representations in Vector Space
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Distributed Representations of Words and Phrases and their Compositionality
• Neural Machine Translation by Jointly Learning to Align and Translate
• Sequence to Sequence Learning with Neural Networks
• ADADELTA: An Adaptive Learning Rate Method
• Distributed Representations of Sentences and Documents
• On the difficulty of training Recurrent Neural Networks
• How transferable are features in deep neural networks?
• Instance Normalization: The Missing Ingredient for Fast Stylization
• Scalable and accurate deep learning for electronic health records
• Deep Learning for Sensor-based Activity Recognition: A Survey
• Imaging Time-Series to Improve Classification and Imputation
• Multi-Scale Convolutional Neural Networks for Time Series Classification
• Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning
• Transfer learning for time series classification
• Generalized Denoising Auto-Encoders as Generative Models
• Detecting and interpreting myocardial infarction using fully convolutional neural networks
• Deep Learning for Time-Series Analysis
• Should we really use post-hoc tests based on mean-ranks?
• Evaluating surgical skills from kinematic data using convolutional neural networks
• TimeNet: Pre-trained deep recurrent neural network for time series classification
• Classification of Time-Series Images Using Deep Convolutional Neural Networks
• Data augmentation using synthetic data for time series classification with deep residual networks
• Deep Echo State Network (DeepESN): A Brief Survey
• Spatio-Temporal Neural Networks for Space-Time Series Forecasting and Relations Discovery
• Time-series modeling with undecimated fully convolutional neural networks
• Spatially Encoding Temporal Correlations to Classify Temporal Data Using Convolutional Neural Networks
• Earliness-Aware Deep Convolutional Networks for Early Time Series Classification
• Cost-Sensitive Convolution based Neural Networks for Imbalanced Time-Series Classification
• Genesis of Basic and Multi-Layer Echo State Network Recurrent Autoencoders for Efficient Data Representations
• Representation Learning with Deconvolution for Multivariate Time Series Classification and Visualization

Cited by in corpus (255)

• Deep Learning for Anomaly Detection: A Survey
• ROCKET: Exceptionally fast and accurate time series classification using random convolutional kernels
• An Experimental Review on Deep Learning Architectures for Time Series Forecasting
• Benchmarking Multivariate Time Series Classification Algorithms
• MINIROCKET: A Very Fast (Almost) Deterministic Transform for Time Series Classification
• ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions
• Self-attention for raw optical Satellite Time Series Classification
• Foundations and modelling of dynamic networks using Dynamic Graph Neural Networks: A survey
• HIVE-COTE 2.0: a new meta ensemble for time series classification
• Transfer learning for time series classification
• Improving Position Encoding of Transformers for Multivariate Time Series Classification
• Unsupervised Scalable Representation Learning for Multivariate Time Series
• TS-CHIEF: A Scalable and Accurate Forest Algorithm for Time Series Classification
• sktime: A Unified Interface for Machine Learning with Time Series
• Bake off redux: a review and experimental evaluation of recent time series classification algorithms
• TAnoGAN: Time Series Anomaly Detection with Generative Adversarial Networks
• Adversarial Attacks on Deep Neural Networks for Time Series Classification
• CARLA: Self-supervised Contrastive Representation Learning for Time Series Anomaly Detection
• A Compact and Interpretable Convolutional Neural Network for Cross-Subject Driver Drowsiness Detection from Single-Channel EEG
• Fully Learnable Deep Wavelet Transform for Unsupervised Monitoring of High-Frequency Time Series
• The Canonical Interval Forest (CIF) Classifier for Time Series Classification
• Evaluation of post-hoc interpretability methods in time-series classification
• Satellite Image Time Series Analysis for Big Earth Observation Data
• Approaches and Applications of Early Classification of Time Series: A Review
• Gated Transformer Networks for Multivariate Time Series Classification
• GRATIS: GeneRAting TIme Series with diverse and controllable characteristics
• Deep Neural Network Ensembles for Time Series Classification
• Computational Solar Energy -- Ensemble Learning Methods for Prediction of Solar Power Generation based on Meteorological Parameters in Eastern India
• Self-supervised Autoregressive Domain Adaptation for Time Series Data
• Classification of chaotic time series with deep learning
• Time-Series Pattern Recognition in Smart Manufacturing Systems: A Literature Review and Ontology
• A machine learning approach for forecasting hierarchical time series
• ADATIME: A Benchmarking Suite for Domain Adaptation on Time Series Data
• Unsupervised Deep Learning for IoT Time Series
• Time-Series Classification in Smart Manufacturing Systems: An Experimental Evaluation of State-of-the-Art Machine Learning Algorithms
• Stochastic Physics-Informed Neural Ordinary Differential Equations
• Deep Learning-based Cattle Activity Classification Using Joint Time-frequency Data Representation
• Ensembles of Localised Models for Time Series Forecasting
• TCGAN: Convolutional Generative Adversarial Network for Time Series Classification and Clustering
• A Spatiotemporal Deep Neural Network for Fine-Grained Multi-Horizon Wind Prediction
• Automatic classification of plasma regions in near-Earth space with supervised machine learning: application to Magnetospheric Multi Scale 2016-2019 observations
• Untargeted, Targeted and Universal Adversarial Attacks and Defenses on Time Series
• Temporal signals to images: Monitoring the condition of industrial assets with deep learning image processing algorithms
• Process Outcome Prediction: CNN vs. LSTM (with Attention)
• Label-efficient Time Series Representation Learning: A Review
• The FreshPRINCE: A Simple Transformation Based Pipeline Time Series Classifier
• Learning future terrorist targets through temporal meta-graphs
• Resource-aware Time Series Imaging Classification for Wireless Link Layer Anomalies
• A Review of 1D Convolutional Neural Networks toward Unknown Substance Identification in Portable Raman Spectrometer
• A comparative study of different machine learning methods for dissipative quantum dynamics
• Random Dilated Shapelet Transform: A New Approach for Time Series Shapelets
• Unsupervised Representation Learning for Time Series with Temporal Neighborhood Coding
• Scoring and Assessment in Medical VR Training Simulators with Dynamic Time Series Classification
• Determination of Trace Organic Contaminant Concentration via Machine Classification of Surface-Enhanced Raman Spectra
• Interpretable Detection of Partial Discharge in Power Lines with Deep Learning
• Adaptive Modeling of Satellite-Derived Nighttime Lights Time-Series for Tracking Urban Change Processes Using Machine Learning
• DeepVATS: Deep Visual Analytics for Time Series
• DACAD: Domain Adaptation Contrastive Learning for Anomaly Detection in Multivariate Time Series
• An Evaluation of Machine Learning and Deep Learning Models for Drought Prediction using Weather Data
• Experiment data-driven modeling of tokamak discharge in EAST
• Similarity Learning based Few Shot Learning for ECG Time Series Classification
• Forecasting the abnormal events at well drilling with machine learning
• dCAM: Dimension-wise Class Activation Map for Explaining Multivariate Data Series Classification
• Deep Volumetric Ambient Occlusion
• Automated scoring of pre-REM sleep in mice with deep learning
• Deep Learning for Cardiologist-level Myocardial Infarction Detection in Electrocardiograms
• Privacy Meets Explainability: A Comprehensive Impact Benchmark
• Time-Series Forecasting in Smart Manufacturing Systems: An Experimental Evaluation of the State-of-the-art Algorithms
• Role of Data Augmentation Strategies in Knowledge Distillation for Wearable Sensor Data
• Low Dimensional Convolutional Neural Network For Solar Flares GOES Time Series Classification
• Explaining Deep Classification of Time-Series Data with Learned Prototypes
• Artificial neural network subgrid models of 2-D compressible magnetohydrodynamic turbulence
• Knock Detection in Combustion Engine Time Series Using a Theory-Guided 1D Convolutional Neural Network Approach
• Unsupervised Feature Based Algorithms for Time Series Extrinsic Regression
• Time series classification for varying length series
• Explainable Machine Learning for Breakdown Prediction in High Gradient RF Cavities
• Generalised gravitational burst generation with Generative Adversarial Networks
• Real-Time Detection of Unmodelled Gravitational-Wave Transients Using Convolutional Neural Networks
• Appliance Detection Using Very Low-Frequency Smart Meter Time Series
• An empirical comparison of deep-neural-network architectures for next activity prediction using context-enriched process event logs
• CycleSense: Detecting Near Miss Incidents in Bicycle Traffic from Mobile Motion Sensors
• Exploring Semi-Supervised Learning for Predicting Listener Backchannels
• Deep learning models for predictive maintenance: a survey, comparison, challenges and prospect
• Adjusting for Autocorrelated Errors in Neural Networks for Time Series
• PyLightcurve-torch: a transit modelling package for deep learning applications in PyTorch
• Causal Inference for Time series Analysis: Problems, Methods and Evaluation
• Astroconformer: The Prospects of Analyzing Stellar Light Curves with Transformer-Based Deep Learning Models
• Robust Explainer Recommendation for Time Series Classification
• Voice2Series: Reprogramming Acoustic Models for Time Series Classification
• AI-driven Java Performance Testing: Balancing Result Quality with Testing Time
• Predicting the Stability of Hierarchical Triple Systems with Convolutional Neural Networks
• Predicting Terrorist Attacks in the United States using Localized News Data
• Surface Similarity Parameter: A New Machine Learning Loss Metric for Oscillatory Spatio-Temporal Data
• In-situ animal behavior classification using knowledge distillation and fixed-point quantization
• Towards an automated data cleaning with deep learning in CRESST
• Model-assisted deep learning of rare extreme events from partial observations
• Orchestrating the Development Lifecycle of Machine Learning-Based IoT Applications: A Taxonomy and Survey
• Detecting subtle cyberattacks on adaptive cruise control vehicles: A machine learning approach
• The Any Light Particle Search experiment at DESY
• My(o) Armband Leaks Passwords: An EMG and IMU Based Keylogging Side-Channel Attack
• Predictive Auto-scaling with OpenStack Monasca
• Benefit-aware Early Prediction of Health Outcomes on Multivariate EEG Time Series
• Probing nuclear physics with supernova gravitational waves and machine learning
• Time Series Featurization via Topological Data Analysis
• DeConFuse : A Deep Convolutional Transform based Unsupervised Fusion Framework
• A Performance Analysis of Basin Hopping Compared to Established Metaheuristics for Global Optimization
• Generalized Label Enhancement with Sample Correlations
• Deep Learning for Human Locomotion Analysis in Lower-Limb Exoskeletons: A Comparative Study
• Sequence-to-Sequence Forecasting-aided State Estimation for Power Systems
• Application of LSTM architectures for next frame forecasting in Sentinel-1 images time series
• Tropical time series, iterated-sums signatures and quasisymmetric functions
• timeXplain -- A Framework for Explaining the Predictions of Time Series Classifiers
• Deep learning for time series classification
• Exploring Kolmogorov-Arnold Networks for Interpretable Time Series Classification
• ALGAN: Time Series Anomaly Detection with Adjusted-LSTM GAN
• Surface Type Classification for Autonomous Robot Indoor Navigation
• Subspace Preserving Quantum Convolutional Neural Network Architectures
• Benchmarking time series classification -- Functional data vs machine learning approaches
• Learning Interpretable Shapelets for Time Series Classification through Adversarial Regularization
• Ask The Machine: Systematic detection of wind-type outflows in low-mass X-ray binaries
• XAI Methods for Neural Time Series Classification: A Brief Review
• Spikebench: An open benchmark for spike train time-series classification
• Automatic Feature Engineering for Time Series Classification: Evaluation and Discussion
• Timage -- A Robust Time Series Classification Pipeline
• Learning Position From Vehicle Vibration Using an Inertial Measurement Unit
• LeaFi: Data Series Indexes on Steroids with Learned Filters
• DTW-Merge: A Novel Data Augmentation Technique for Time Series Classification
• A first application of machine and deep learning for background rejection in the ALPS II TES detector
• Sensitivity - Local Index to Control Chaoticity or Gradient Globally -
• Denoising Architecture for Unsupervised Anomaly Detection in Time-Series
• Deep Semi-Supervised Learning for Time Series Classification
• Deep Active Audio Feature Learning in Resource-Constrained Environments
• Innovative Platform for Designing Hybrid Collaborative & Context-Aware Data Mining Scenarios
• Condition Assessment of Stay Cables through Enhanced Time Series Classification Using a Deep Learning Approach
• Evolutionary Training and Abstraction Yields Algorithmic Generalization of Neural Computers
• RTFN: A Robust Temporal Feature Network for Time Series Classification
• Bridging the Gap: Gaze Events as Interpretable Concepts to Explain Deep Neural Sequence Models
• Multi-task CNN Behavioral Embedding Model For Transaction Fraud Detection
• EavesDroid: Eavesdropping User Behaviors via OS Side-Channels on Smartphones
• FedST: Secure Federated Shapelet Transformation for Time Series Classification
• Machine learning for the diagnosis of early stage diabetes using temporal glucose profiles
• Feature Selection for Multivariate Time Series via Network Pruning
• cDVGAN: One Flexible Model for Multi-class Gravitational Wave Signal and Glitch Generation
• Radar Emitter Classification with Attribute-specific Recurrent Neural Networks
• Swarm Characteristics Classification Using Neural Networks
• Can we Estimate Truck Accident Risk from Telemetric Data using Machine Learning?
• ADF & TransApp: A Transformer-Based Framework for Appliance Detection Using Smart Meter Consumption Series
• How neural networks learn to classify chaotic time series
• Leveraging Deep Learning for Time Series Extrinsic Regression in predicting photometric metallicity of Fundamental-mode RR Lyrae Stars
• Boosting Photon-Number-Resolved Detection Rates of Transition-Edge Sensors by Machine Learning
• Explainable Multivariate Time Series Classification: A Deep Neural Network Which Learns To Attend To Important Variables As Well As Informative Time Intervals
• Predicting Short Term Energy Demand in Smart Grid: A Deep Learning Approach for Integrating Renewable Energy Sources in Line with SDGs 7, 9, and 13
• Seq2Tens: An Efficient Representation of Sequences by Low-Rank Tensor Projections
• Attention module improves both performance and interpretability of 4D fMRI decoding neural network
• Confidence-Guided Learning Process for Continuous Classification of Time Series
• Hardware-Friendly Delayed-Feedback Reservoir for Multivariate Time-Series Classification
• Low Cost Carriers induce specific and identifiable delay propagation patterns: an analysis of the EU and US systems
• Unsupervised Visual Time-Series Representation Learning and Clustering
• Merging Subject Matter Expertise and Deep Convolutional Neural Network for State-Based Online Machine-Part Interaction Classification
• Interpretable Feature Construction for Time Series Extrinsic Regression
• A Hybrid Deep Learning Model for Predictive Flood Warning and Situation Awareness using Channel Network Sensors Data
• Ground Reaction Force Estimation via Time-aware Knowledge Distillation
• Pattern Discovery in Time Series with Byte Pair Encoding
• Animal Behavior Classification via Accelerometry Data and Recurrent Neural Networks
• DEBACER: a method for slicing moderated debates
• A non-autonomous equation discovery method for time signal classification
• Merchant Category Identification Using Credit Card Transactions
• Consumer Behaviour in Retail: Next Logical Purchase using Deep Neural Network
• Persistent homology of featured time series data and its applications
• Wearable data from subjects playing Super Mario, sitting university exams, or performing physical exercise help detect acute mood episodes via self-supervised learning
• Detecting Pulsars with Neural Networks: A Proof of Concept
• Time series analysis using persistent homology of distance matrix
• Interpretable Time-series Classification on Few-shot Samples
• Fingerprint Recognition of Partial Discharge Signals in Deep Learning Enhanced Rydberg Atomic Sensors
• DTWSSE: Data Augmentation with a Siamese Encoder for Time Series
• Deep Unsupervised Domain Adaptation for Time Series Classification: a Benchmark
• Mobility-based prediction of SARS-CoV-2 spreading
• Short-Term Stock Price-Trend Prediction Using Meta-Learning
• ATCN: Resource-Efficient Processing of Time Series on Edge
• DeepHeteroIoT: Deep Local and Global Learning over Heterogeneous IoT Sensor Data
• Towards Diverse Perspective Learning with Selection over Multiple Temporal Poolings
• Nyström Regularization for Time Series Forecasting
• IaC Generation with LLMs: An Error Taxonomy and A Study on Configuration Knowledge Injection
• Generative Pretrained Embedding and Hierarchical Irregular Time Series Representation for Daily Living Activity Recognition
• Behavioral Biometrics for Automatic Detection of User Familiarity in VR
• Minimum variance threshold for epsilon-lexicase selection
• FlexParser -- the adaptive log file parser for continuous results in a changing world
• An End-To-End Analysis of Deep Learning-Based Remaining Useful Life Algorithms for Satefy-Critical 5G-Enabled IIoT Networks
• Diffeomorphic Transformations for Time Series Analysis: An Efficient Approach to Nonlinear Warping
• A Multi-label Time Series Classification Approach for Non-intrusive Water End-Use Monitoring
• RTFN: Robust Temporal Feature Network
• Adaptive Law-Based Transformation (ALT): A Lightweight Feature Representation for Time Series Classification
• TE-ESN: Time Encoding Echo State Network for Prediction Based on Irregularly Sampled Time Series Data
• An Artificial Intelligence Solution for Electricity Procurement in Forward Markets
• Sketches for Time-Dependent Machine Learning
• Modeling Latent Non-Linear Dynamical System over Time Series
• ForecastQA: A Question Answering Challenge for Event Forecasting with Temporal Text Data
• Detection of cellular micromotion by advanced signal processing
• RED CoMETS: An ensemble classifier for symbolically represented multivariate time series
• DYSAN: Dynamically sanitizing motion sensor data against sensitive inferences through adversarial networks
• MEME: Generating RNN Model Explanations via Model Extraction
• Attention to Warp: Deep Metric Learning for Multivariate Time Series
• Depth Evaluation for Metal Surface Defects by Eddy Current Testing using Deep Residual Convolutional Neural Networks
• Learnable Dynamic Temporal Pooling for Time Series Classification
• Fitbeat: COVID-19 Estimation based on Wristband Heart Rate
• Class-Dependent Perturbation Effects in Evaluating Time Series Attributions
• Conspiracy to Commit: Information Pollution, Artificial Intelligence, and Real-World Hate Crime
• TEASER: Early and Accurate Time Series Classification
• Driver Identification through Stochastic Multi-State Car-Following Modeling
• Leveraging the Self-Transition Probability of Ordinal Pattern Transition Graph for Transportation Mode Classification
• Neural models for prediction of spatially patterned phase transitions: methods and challenges
• Few Labels are all you need: A Weakly Supervised Framework for Appliance Localization in Smart-Meter Series
• Collaborative Inference for Efficient Remote Monitoring
• Active Transfer Prototypical Network: An Efficient Labeling Algorithm for Time-Series Data
• Detecting Driver's Distraction using Long-term Recurrent Convolutional Network
• Deep Learning for Time Series Classification of Parkinson's Disease Eye Tracking Data
• Uncertainty-Aware Multiple Instance Learning from Large-Scale Long Time Series Data
• Identity Recognition in Intelligent Cars with Behavioral Data and LSTM-ResNet Classifier
• Deep Joint Distribution Optimal Transport for Universal Domain Adaptation on Time Series
• Modeling Irregular Astronomical Time Series with Neural Stochastic Delay Differential Equations
• Estimating the electrical power output of industrial devices with end-to-end time-series classification in the presence of label noise
• Mid infrared spectroscopy and milk quality traits: a data analysis competition at the "International Workshop on Spectroscopy and Chemometrics 2021"
• Enhancing Time Series Classification with Diversity-Driven Neural Network Ensembles
• Green-NAS: A Global-Scale Multi-Objective Neural Architecture Search for Robust and Efficient Edge-Native Weather Forecasting
• Long-Term Dynamical Evolution and Ejection of Near-Earth Asteroids
• Enhancing deep learning models for time series classification via knowledge distillation
• Accuracy Improvement for Fully Convolutional Networks via Selective Augmentation with Applications to Electrocardiogram Data
• A systematic review of Python packages for time series analysis
• Selective Correlation Based Knowledge Distillation for Ground Reaction Force Estimation
• Neural networks in the search for fast radio bursts with RATAN-600
• Online Training and Inference System on Edge FPGA Using Delayed Feedback Reservoir
• Classification of multivariate weakly-labelled time-series with attention
• Neural Networks for the Analysis of Traced Particles in Kinetic Plasma Simulations
• Boosting Certified Robustness for Time Series Classification with Efficient Self-Ensemble
• Predictive maintenance on event logs: Application on an ATM fleet
• Contralaterally Enhanced Networks for Thoracic Disease Detection
• Piece-wise Matching Layer in Representation Learning for ECG Classification
• Fast, Accurate and Interpretable Time Series Classification Through Randomization
• GeoStat Representations of Time Series for Fast Classification
• Improving Time Series Classification Algorithms Using Octave-Convolutional Layers
• Coalitional Bayesian Autoencoders -- Towards explainable unsupervised deep learning
• Light curve fingerprints: an automated approach to the extraction of X-ray variability patterns with feature aggregation -- an example application to GRS 1915+105
• Identification of the nature of dynamical systems with recurrence plots and convolution neural networks: A preliminary test
• Deep Neural Networks on EEG Signals to Predict Auditory Attention Score Using Gramian Angular Difference Field
• On tuning deep learning models: a data mining perspective
• KryptoOracle: A Real-Time Cryptocurrency Price Prediction Platform Using Twitter Sentiments
• Anomaly Detection in Time Series with Triadic Motif Fields and Application in Atrial Fibrillation ECG Classification
• Multi-Faceted Representation Learning with Hybrid Architecture for Time Series Classification
• ODE guided Neural Data Augmentation Techniques for Time Series Data and its Benefits on Robustness
• Benchmarking Hybrid Deep Learning Architectures for Predictive Maintenance in Industry 4.0
• Modeling continuous-time stochastic processes using $\mathcal{N}$-Curve mixtures
• Co-eye: A Multi-resolution Symbolic Representation to TimeSeries Diversified Ensemble Classification
• Gated Res2Net for Multivariate Time Series Analysis
• Dynamic Time Warp Convolutional Networks
• Analysis of Gait-Event-related Brain Potentials During Instructed And Spontaneous Treadmill Walking -- Technical Affordances and used Methods