Time-series generated by end-users, edge devices, and different wearables are mostly unlabelled. We propose a method to auto-generate labels of un-labelled time-series, exploiting very few representative labelled time-series. Our method is based on representation learning using Auto Encoded Compact Sequence (AECS) with a choice of best distance measure. It performs self-correction in iterations, by learning latent structure, as well as synthetically boosting representative time-series using Variational-Auto-Encoder (VAE) to improve the quality of labels. We have experimented with UCR and UCI archives, public real-world univariate, multivariate time-series taken from different application domains. Experimental results demonstrate that the proposed method is very close to the performance achieved by fully supervised classification. The proposed method not only produces close to benchmark results but outperforms the benchmark performance in some cases.
Spike-based neuromorphic hardware holds the promise to provide more energy efficient implementations of Deep Neural Networks (DNNs) than standard hardware such as GPUs. But this requires to understand how DNNs can be emulated in an event-based sparse firing regime, since otherwise the energy-advantage gets lost. In particular, DNNs that solve sequence processing tasks typically employ Long Short-Term Memory (LSTM) units that are hard to emulate with few spikes. We show that a facet of many biological neurons, slow after-hyperpolarizing (AHP) currents after each spike, provides an efficient solution. AHP-currents can easily be implemented in neuromorphic hardware that supports multi-compartment neuron models, such as Intel's Loihi chip. Filter approximation theory explains why AHP-neurons can emulate the function of LSTM units. This yields a highly energy-efficient approach to time series classification. Furthermore it provides the basis for implementing with very sparse firing an important class of large DNNs that extract relations between words and sentences in a text in order to answer questions about the text.
Motor disturbances can affect the interaction with dynamic objects, such as catching a ball. A classification of clinical catching trials might give insight into the existence of pathological alterations in the relation of arm and ball movements. Accurate, but also early decisions are required to classify a catching attempt before the catcher's first ball contact. To obtain clinically valuable results, a significant decision confidence of at least 75% is required. Hence, three competing objectives have to be optimized at the same time: accuracy, earliness and decision-making confidence. Here we propose a coupled classification and prediction approach for early time series classification: a predictive, generative recurrent neural network (RNN) forecasts the next data points of ball trajectories based on already available observations; a discriminative RNN continuously generates classification guesses based on the available data points and the unrolled sequence predictions. We compare our approach, which we refer to as predictive sequential classification (PSC), to state-of-the-art sequence learners, including various RNN and temporal convolutional network (TCN) architectures. On this hard real-world task we can consistently demonstrate the superiority of PSC over all other models in terms of accuracy and confidence with respect to earliness of recognition. Specifically, PSC is able to confidently classify the success of catching trials as early as 123 milliseconds before the first ball contact. We conclude that PSC is a promising approach for early time series classification, when accurate and confident decisions are required.
High levels of missing data and strong class imbalance are ubiquitous challenges that are often presented simultaneously in real-world time series data. Existing methods approach these problems separately, frequently making significant assumptions about the underlying data generation process in order to lessen the impact of missing information. In this work, we instead demonstrate how a general self-supervised training method, namely Autoregressive Predictive Coding (APC), can be leveraged to overcome both missing data and class imbalance simultaneously without strong assumptions. Specifically, on a synthetic dataset, we show that standard baselines are substantially improved upon through the use of APC, yielding the greatest gains in the combined setting of high missingness and severe class imbalance. We further apply APC on two real-world medical time-series datasets, and show that APC improves the classification performance in all settings, ultimately achieving state-of-the-art AUPRC results on the Physionet benchmark.
Learning decent representations from unlabeled time-series data with temporal dynamics is a very challenging task. In this paper, we propose an unsupervised Time-Series representation learning framework via Temporal and Contextual Contrasting (TS-TCC), to learn time-series representation from unlabeled data. First, the raw time-series data are transformed into two different yet correlated views by using weak and strong augmentations. Second, we propose a novel temporal contrasting module to learn robust temporal representations by designing a tough cross-view prediction task. Last, to further learn discriminative representations, we propose a contextual contrasting module built upon the contexts from the temporal contrasting module. It attempts to maximize the similarity among different contexts of the same sample while minimizing similarity among contexts of different samples. Experiments have been carried out on three real-world time-series datasets. The results manifest that training a linear classifier on top of the features learned by our proposed TS-TCC performs comparably with the supervised training. Additionally, our proposed TS-TCC shows high efficiency in few-labeled data and transfer learning scenarios. The code is publicly available at https://github.com/emadeldeen24/TS-TCC.
This paper presents TS2Vec, a universal framework for learning representations of time series in an arbitrary semantic level. Unlike existing methods, TS2Vec performs contrastive learning in a hierarchical way over augmented context views, which enables a robust contextual representation for each timestamp. Furthermore, to obtain the representation of an arbitrary sub-sequence in the time series, we can apply a simple aggregation over the representations of corresponding timestamps. We conduct extensive experiments on time series classification tasks to evaluate the quality of time series representations. As a result, TS2Vec achieves significant improvement over existing SOTAs of unsupervised time series representation on 125 UCR datasets and 29 UEA datasets. The learned timestamp-level representations also achieve superior results in time series forecasting and anomaly detection tasks. A linear regression trained on top of the learned representations outperforms previous SOTAs of time series forecasting. Furthermore, we present a simple way to apply the learned representations for unsupervised anomaly detection, which establishes SOTA results in the literature. The source code is publicly available at https://github.com/yuezhihan/ts2vec.
Multivariable time series classification problems are increasing in prevalence and complexity in a variety of domains, such as biology and finance. While deep learning methods are an effective tool for these problems, they often lack interpretability. In this work, we propose a novel modular prototype learning framework for multivariable time series classification. In the first stage of our framework, encoders extract features from each variable independently. Prototype layers identify single-variable prototypes in the resulting feature spaces. The next stage of our framework represents the multivariable time series sample points in terms of their similarity to these single-variable prototypes. This results in an inherently interpretable representation of multivariable patterns, on which prototype learning is applied to extract representative examples i.e. multivariable prototypes. Our framework is thus able to explicitly identify both informative patterns in the individual variables, as well as the relationships between the variables. We validate our framework on a simulated dataset with embedded patterns, as well as a real human activity recognition problem. Our framework attains comparable or superior classification performance to existing time series classification methods on these tasks. On the simulated dataset, we find that our model returns interpretations consistent with the embedded patterns. Moreover, the interpretations learned on the activity recognition dataset align with domain knowledge.
Learning to classify time series with limited data is a practical yet challenging problem. Current methods are primarily based on hand-designed feature extraction rules or domain-specific data augmentation. Motivated by the advances in deep speech processing models and the fact that voice data are univariate temporal signals, in this paper, we propose Voice2Series (V2S), a novel end-to-end approach that reprograms acoustic models for time series classification, through input transformation learning and output label mapping. Leveraging the representation learning power of a large-scale pre-trained speech processing model, on 30 different time series tasks we show that V2S performs competitive results on 19 time series classification tasks. We further provide a theoretical justification of V2S by proving its population risk is upper bounded by the source risk and a Wasserstein distance accounting for feature alignment via reprogramming. Our results offer new and effective means to time series classification.
The amount of information produced, whether by newspapers, blogs and social networks, or by monitoring systems, is increasing rapidly. Processing all this data in real-time, while taking into consideration advanced knowledge about the problem domain, is challenging, but required in scenarios where assessing potential risks in a timely fashion is critical. C-SPARQL, a language for continuous queries over streams of RDF data, is one of the more prominent approaches in stream reasoning that provides such continuous inference capabilities over dynamic data that go beyond mere stream processing. However, it has been shown that, in the presence of huge amounts of data, C-SPARQL may not be able to answer queries in time, in particular when the frequency of incoming data is higher than the time required for reasoning with that data. In this paper, we investigate whether reasoning with C-SPARQL can be approximated using Recurrent Neural Networks and Convolutional Neural Networks, two neural network architectures that have been shown to be well-suited for time series forecasting and time series classification, to leverage on their higher processing speed once the network has been trained. We consider a variety of different kinds of queries and obtain overall positive results with high accuracies while improving processing time often by several orders of magnitude.
We detail a novel class of implicit neural models. Leveraging time-parallel methods for differential equations, Multiple Shooting Layers (MSLs) seek solutions of initial value problems via parallelizable root-finding algorithms. MSLs broadly serve as drop-in replacements for neural ordinary differential equations (Neural ODEs) with improved efficiency in number of function evaluations (NFEs) and wall-clock inference time. We develop the algorithmic framework of MSLs, analyzing the different choices of solution methods from a theoretical and computational perspective. MSLs are showcased in long horizon optimal control of ODEs and PDEs and as latent models for sequence generation. Finally, we investigate the speedups obtained through application of MSL inference in neural controlled differential equations (Neural CDEs) for time series classification of medical data.
Embedding nonlinear dynamical systems into artificial neural networks is a powerful new formalism for machine learning. By parameterizing ordinary differential equations (ODEs) as neural network layers, these Neural ODEs are memory-efficient to train, process time-series naturally and incorporate knowledge of physical systems into deep learning models. However, the practical applications of Neural ODEs are limited due to long inference times, because the outputs of the embedded ODE layers are computed numerically with differential equation solvers that can be computationally demanding. Here we show that mathematical model order reduction methods can be used for compressing and accelerating Neural ODEs by accurately simulating the continuous nonlinear dynamics in low-dimensional subspaces. We implement our novel compression method by developing Neural ODEs that integrate the necessary subspace-projection and interpolation operations as layers of the neural network. We validate our approach by comparing it to neuron pruning and SVD-based weight truncation methods from the literature in image and time-series classification tasks. The methods are evaluated by acceleration versus accuracy when adjusting the level of compression. On this spectrum, we achieve a favourable balance over existing methods by using model order reduction when compressing a convolutional Neural ODE. In compressing a recurrent Neural ODE, SVD-based weight truncation yields good performance. Based on our results, our integration of model order reduction with Neural ODEs can facilitate efficient, dynamical system-driven deep learning in resource-constrained applications.
Time series classification (TSC) aims to predict the class label of a given time series, which is critical to a rich set of application areas such as economics and medicine. State-of-the-art TSC methods have mostly focused on classification accuracy and efficiency, without considering the interpretability of their classifications, which is an important property required by modern applications such as appliance modeling and legislation such as the European General Data Protection Regulation. To address this gap, we propose a novel TSC method - the Randomized-Supervised Time Series Forest (r-STSF). r-STSF is highly efficient, achieves state-of-the-art classification accuracy and enables interpretability. r-STSF takes an efficient interval-based approach to classify time series according to aggregate values of discriminatory sub-series (intervals). To achieve state-of-the-art accuracy, r-STSF builds an ensemble of randomized trees using the discriminatory sub-series. It uses four time series representations, nine aggregation functions and a supervised binary-inspired search combined with a feature ranking metric to identify highly discriminatory sub-series. The discriminatory sub-series enable interpretable classifications. Experiments on extensive datasets show that r-STSF achieves state-of-the-art accuracy while being orders of magnitude faster than most existing TSC methods. It is the only classifier from the state-of-the-art group that enables interpretability. Our findings also highlight that r-STSF is the best TSC method when classifying complex time series datasets.
While the big data revolution takes place, large amounts of electronic health records, such as electrocardiograms (ECGs) and vital signs data, have become available. These signals are often recorded as time series of observations and are now easier to obtain. In particular, with the arise of smart devices that can perform ECG, there is the quest for developing novel approaches that allow to monitor these signals efficiently, and quickly detect anomalies. However, since most data generated remains unlabelled, the task of anomaly detection is still very challenging. Unsupervised representation learning using deep generative models (e.g., variational autoencoders) has been used to learn expressive feature representations of sequences that can make downstream tasks, such as anomaly detection, easier to execute and more accurate. We propose an approach for unsupervised representation learning of ECG sequences using a variational autoencoder parameterised by recurrent neural networks, and use the learned representations for anomaly detection using multiple detection strategies. We tested our approach on the ECG5000 electrocardiogram dataset of the UCR time series classification archive. Our results show that the proposed approach is able to learn expressive representations of ECG sequences, and to detect anomalies with scores that outperform other both supervised and unsupervised methods.
The research in anomaly detection lacks a unified definition of what represents an anomalous instance. Discrepancies in the nature itself of an anomaly lead to multiple paradigms of algorithms design and experimentation. Predictive maintenance is a special case, where the anomaly represents a failure that must be prevented. Related time-series research as outlier and novelty detection or time-series classification does not apply to the concept of an anomaly in this field, because they are not single points which have not been seen previously and may not be precisely annotated. Moreover, due to the lack of annotated anomalous data, many benchmarks are adapted from supervised scenarios. To address these issues, we generalise the concept of positive and negative instances to intervals to be able to evaluate unsupervised anomaly detection algorithms. We also preserve the imbalance scheme for evaluation through the proposal of the Preceding Window ROC, a generalisation for the calculation of ROC curves for time-series scenarios. We also adapt the mechanism from a established time-series anomaly detection benchmark to the proposed generalisations to reward early detection. Therefore, the proposal represents a flexible evaluation framework for the different scenarios. To show the usefulness of this definition, we include a case study of Big Data algorithms with a real-world time-series problem provided by the company ArcelorMittal, and compare the proposal with an evaluation method.
Future surveys such as the Legacy Survey of Space and Time (LSST) of the Vera C. Rubin Observatory will observe an order of magnitude more astrophysical transient events than any previous survey before. With this deluge of photometric data, it will be impossible for all such events to be classified by humans alone. Recent efforts have sought to leverage machine learning methods to tackle the challenge of astronomical transient classification, with ever improving success. Transformers are a recently developed deep learning architecture, first proposed for natural language processing, that have shown a great deal of recent success. In this work we develop a new transformer architecture, which uses multi-head self attention at its core, for general multi-variate time-series data. Furthermore, the proposed time-series transformer architecture supports the inclusion of an arbitrary number of additional features, while also offering interpretability. We apply the time-series transformer to the task of photometric classification, minimising the reliance of expert domain knowledge for feature selection, while achieving results comparable to state-of-the-art photometric classification methods. We achieve a logarithmic-loss of 0.507 on imbalanced data in a representative setting using data from the Photometric LSST Astronomical Time-Series Classification Challenge (PLAsTiCC). Moreover, we achieve a micro-averaged receiver operating characteristic area under curve of 0.98 and micro-averaged precision-recall area under curve of 0.87.
Using bag of words representations of time series is a popular approach to time series classification. These algorithms involve approximating and discretising windows over a series to form words, then forming a count of words over a given dictionary. Classifiers are constructed on the resulting histograms of word counts. A 2017 evaluation of a range of time series classifiers found the bag of symbolic-fourier approximation symbols (BOSS) ensemble the best of the dictionary based classifiers. It forms one of the components of hierarchical vote collective of transformation-based ensembles (HIVE-COTE), which represents the current state of the art. Since then, several new dictionary based algorithms have been proposed that are more accurate or more scalable (or both) than BOSS. We propose a further extension of these dictionary based classifiers that combines the best elements of the others combined with a novel approach to constructing ensemble members based on an adaptive Gaussian process model of the parameter space. We demonstrate that the temporal dictionary ensemble (TDE) is more accurate than other dictionary based approaches. Furthermore, unlike the other classifiers, if we replace BOSS in HIVE-COTE with TDE, HIVE-COTE is significantly more accurate. We also show this new version of HIVE-COTE is significantly more accurate than the current best deep learning approach, a recently proposed hybrid tree ensemble and a recently introduced competitive classifier making use of highly randomised convolutional kernels. This advance represents a new state of the art for time series classification.
An all-clear flare prediction is a type of solar flare forecasting that puts more emphasis on predicting non-flaring instances (often relatively small flares and flare quiet regions) with high precision while still maintaining valuable predictive results. While many flare prediction studies do not address this problem directly, all-clear predictions can be useful in operational context. However, in all-clear predictions, finding the right balance between avoiding false negatives (misses) and reducing the false positives (false alarms) is often challenging. Our study focuses on training and testing a set of interval-based time series classifiers named Time Series Forest (TSF). These classifiers will be used towards building an all-clear flare prediction system by utilizing multivariate time series data. Throughout this paper, we demonstrate our data collection, predictive model building and evaluation processes, and compare our time series classification models with baselines using our benchmark datasets. Our results show that time series classifiers provide better forecasting results in terms of skill scores, precision and recall metrics, and they can be further improved for more precise all-clear forecasts by tuning model hyperparameters.
In complex industrial settings, it is common practice to monitor the operation of machines in order to detect undesired states, adjust maintenance schedules, optimize system performance or collect usage statistics of individual machines. In this work, we focus on estimating the power output of a Combined Heat and Power (CHP) machine of a medium-sized company facility by analyzing the total facility power consumption. We formulate the problem as a time-series classification problem where the class label represents the CHP power output. As the facility is fully instrumented and sensor measurements from the CHP are available, we generate the training labels in an automated fashion from the CHP sensor readings. However, sensor failures result in mislabeled training data samples which are hard to detect and remove from the dataset. Therefore, we propose a novel multi-task deep learning approach that jointly trains a classifier and an autoencoder with a shared embedding representation. The proposed approach targets to gradually correct the mislabelled data samples during training in a self-supervised fashion, without any prior assumption on the amount of label noise. We benchmark our approach on several time-series classification datasets and find it to be comparable and sometimes better than state-of-the-art methods. On the real-world use-case of predicting the CHP power output, we thoroughly evaluate the architectural design choices and show that the final architecture considerably increases the robustness of the learning process and consistently beats other recent state-of-the-art algorithms in the presence of unstructured as well as structured label noise.
Visibility Graph (VG) transforms time series into graphs, facilitating signal processing by advanced graph data mining algorithms. In this paper, based on the classic Limited Penetrable Visibility Graph (LPVG) method, we propose a novel nonlinear mapping method named Circular Limited Penetrable Visibility Graph (CLPVG). The testing on degree distribution and clustering coefficient on the generated graphs of typical time series validates that our CLPVG is able to effectively capture the important features of time series and has better anti-noise ability than traditional LPVG. The experiments on real-world time-series datasets of radio signal and electroencephalogram (EEG) also suggest that the structural features provided by CLPVG, rather than LPVG, are more useful for time-series classification, leading to higher accuracy. And this classification performance can be further enhanced through structural feature expansion by adopting Subgraph Networks (SGN). All of these results validate the effectiveness of our CLPVG model.
Deep learning methods have shown suitability for time series classification in the health and medical domain, with promising results for electrocardiogram data classification. Successful identification of myocardial infarction holds life saving potential and any meaningful improvement upon deep learning models in this area is of great interest. Conventionally, data augmentation methods are applied universally to the training set when data are limited in order to ameliorate data resolution or sample size. In the method proposed in this study, data augmentation was not applied in the context of data scarcity. Instead, samples that yield low confidence predictions were selectively augmented in order to bolster the model's sensitivity to features or patterns less strongly associated with a given class. This approach was tested for improving the performance of a Fully Convolutional Network. The proposed approach achieved 90 percent accuracy for classifying myocardial infarction as opposed to 82 percent accuracy for the baseline, a marked improvement. Further, the accuracy of the proposed approach was optimal near a defined upper threshold for qualifying low confidence samples and decreased as this threshold was raised to include higher confidence samples. This suggests exclusively selecting lower confidence samples for data augmentation comes with distinct benefits for electrocardiogram data classification with Fully Convolutional Networks.