Correlated time series analysis plays an important role in many real-world industries. Learning an efficient representation of this large-scale data for further downstream tasks is necessary but challenging. In this paper, we propose a time-step-level representation learning framework for individual instances via bootstrapped spatiotemporal representation prediction. We evaluated the effectiveness and flexibility of our representation learning framework on correlated time series forecasting and cold-start transferring the forecasting model to new instances with limited data. A linear regression model trained on top of the learned representations demonstrates our model performs best in most cases. Especially compared to representation learning models, we reduce the RMSE, MAE, and MAPE by 37%, 49%, and 48% on the PeMS-BAY dataset, respectively. Furthermore, in real-world metro passenger flow data, our framework demonstrates the ability to transfer to infer future information of new cold-start instances, with gains of 15%, 19%, and 18%. The source code will be released under the GitHub https://github.com/bonaldli/Spatiotemporal-TS-Representation-Learning
Various time variant non-stationary signals need to be pre-processed properly in hydrological time series forecasting in real world, for example, predictions of water level. Decomposition method is a good candidate and widely used in such a pre-processing problem. However, decomposition methods with an inappropriate sampling technique may introduce future data which is not available in practical applications, and result in incorrect decomposition-based forecasting models. In this work, a novel Fully Stepwise Decomposition-Based (FSDB) sampling technique is well designed for the decomposition-based forecasting model, strictly avoiding introducing future information. This sampling technique with decomposition methods, such as Variational Mode Decomposition (VMD) and Singular spectrum analysis (SSA), is applied to predict water level time series in three different stations of Guoyang and Chaohu basins in China. Results of VMD-based hybrid model using FSDB sampling technique show that Nash-Sutcliffe Efficiency (NSE) coefficient is increased by 6.4%, 28.8% and 7.0% in three stations respectively, compared with those obtained from the currently most advanced sampling technique. In the meantime, for series of SSA-based experiments, NSE is increased by 3.2%, 3.1% and 1.1% respectively. We conclude that the newly developed FSDB sampling technique can be used to enhance the performance of decomposition-based hybrid model in water level time series forecasting in real world.
Sliding window sums are widely used for string indexing, hashing and time series analysis. We have developed a family of the generic vectorized sliding sum algorithms that provide speedup of O(P/w) for window size $w$ and number of processors P. For a sum with a commutative operator the speedup is improved to O(P/log(w)). Even more important, our algorithms exhibit efficient memory access patterns. In this paper we study the application of the sliding sum algorithms to the training and inference of the Deep Neural Networks. We demonstrate how both pooling and convolution primitives could be expressed as sliding sums and evaluated by the compute kernels with the shared structure. We show that the sliding sum convolution kernels are more efficient than the commonly used GEMM kernels on the CPU, and could even outperform their GPU counterparts.
An important problem in time-series analysis is modeling systems with time-varying dynamics. Probabilistic models with joint continuous and discrete latent states offer interpretable, efficient, and experimentally useful descriptions of such data. Commonly used models include autoregressive hidden Markov models (ARHMMs) and switching linear dynamical systems (SLDSs), each with its own advantages and disadvantages. ARHMMs permit exact inference and easy parameter estimation, but are parameter intensive when modeling long dependencies, and hence are prone to overfitting. In contrast, SLDSs can capture long-range dependencies in a parameter efficient way through Markovian latent dynamics, but present an intractable likelihood and a challenging parameter estimation task. In this paper, we propose switching autoregressive low-rank tensor (SALT) models, which retain the advantages of both approaches while ameliorating the weaknesses. SALT parameterizes the tensor of an ARHMM with a low-rank factorization to control the number of parameters and allow longer range dependencies without overfitting. We prove theoretical and discuss practical connections between SALT, linear dynamical systems, and SLDSs. We empirically demonstrate quantitative advantages of SALT models on a range of simulated and real prediction tasks, including behavioral and neural datasets. Furthermore, the learned low-rank tensor provides novel insights into temporal dependencies within each discrete state.
IoT time series analysis has found numerous applications in a wide variety of areas, ranging from health informatics to network security. Nevertheless, the complex spatial temporal dynamics and high dimensionality of IoT time series make the analysis increasingly challenging. In recent years, the powerful feature extraction and representation learning capabilities of deep learning (DL) have provided an effective means for IoT time series analysis. However, few existing surveys on time series have systematically discussed unsupervised DL-based methods. To fill this void, we investigate unsupervised deep learning for IoT time series, i.e., unsupervised anomaly detection and clustering, under a unified framework. We also discuss the application scenarios, public datasets, existing challenges, and future research directions in this area.
We consider the problem of heteroscedastic linear regression, where, given $n$ samples $(\mathbf{x}_i, y_i)$ from $y_i = \langle \mathbf{w}^{*}, \mathbf{x}_i \rangle + \epsilon_i \cdot \langle \mathbf{f}^{*}, \mathbf{x}_i \rangle$ with $\mathbf{x}_i \sim N(0,\mathbf{I})$, $\epsilon_i \sim N(0,1)$, we aim to estimate $\mathbf{w}^{*}$. Beyond classical applications of such models in statistics, econometrics, time series analysis etc., it is also particularly relevant in machine learning when data is collected from multiple sources of varying but apriori unknown quality. Our work shows that we can estimate $\mathbf{w}^{*}$ in squared norm up to an error of $\tilde{O}\left(\|\mathbf{f}^{*}\|^2 \cdot \left(\frac{1}{n} + \left(\frac{d}{n}\right)^2\right)\right)$ and prove a matching lower bound (upto log factors). This represents a substantial improvement upon the previous best known upper bound of $\tilde{O}\left(\|\mathbf{f}^{*}\|^2\cdot \frac{d}{n}\right)$. Our algorithm is an alternating minimization procedure with two key subroutines 1. An adaptation of the classical weighted least squares heuristic to estimate $\mathbf{w}^{*}$, for which we provide the first non-asymptotic guarantee. 2. A nonconvex pseudogradient descent procedure for estimating $\mathbf{f}^{*}$ inspired by phase retrieval. As corollaries, we obtain fast non-asymptotic rates for two important problems, linear regression with multiplicative noise and phase retrieval with multiplicative noise, both of which are of independent interest. Beyond this, the proof of our lower bound, which involves a novel adaptation of LeCam's method for handling infinite mutual information quantities (thereby preventing a direct application of standard techniques like Fano's method), could also be of broader interest for establishing lower bounds for other heteroscedastic or heavy-tailed statistical problems.
The modeling of high-frequency data that qualify financial asset transactions has been an area of relevant interest among statisticians and econometricians -- above all, the analysis of time series of financial durations. Autoregressive conditional duration (ACD) models have been the main tool for modeling financial transaction data, where duration is usually defined as the time interval between two successive events. These models are usually specified in terms of a time-varying mean (or median) conditional duration. In this paper, a new extension of ACD models is proposed which is built on the basis of log-symmetric distributions reparametrized by their quantile. The proposed quantile log-symmetric conditional duration autoregressive model allows us to model different percentiles instead of the traditionally used conditional mean (or median) duration. We carry out an in-depth study of theoretical properties and practical issues, such as parameter estimation using maximum likelihood method and diagnostic analysis based on residuals. A detailed Monte Carlo simulation study is also carried out to evaluate the performance of the proposed models and estimation method in retrieving the true parameter values as well as to evaluate a form of residuals. Finally, the proposed class of models is applied to a price duration data set and then used to derive a semi-parametric intraday value-at-risk (IVaR) model.
Traditional hidden Markov models have been a useful tool to understand and model stochastic dynamic linear data; in the case of non-Gaussian data or not linear in mean data, models such as mixture of Gaussian hidden Markov models suffer from the computation of precision matrices and have a lot of unnecessary parameters. As a consequence, such models often perform better when it is assumed that all variables are independent, a hypothesis that may be unrealistic. Hidden Markov models based on kernel density estimation is also capable of modeling non Gaussian data, but they assume independence between variables. In this article, we introduce a new hidden Markov model based on kernel density estimation, which is capable of introducing kernel dependencies using context-specific Bayesian networks. The proposed model is described, together with a learning algorithm based on the expectation-maximization algorithm. Additionally, the model is compared with related HMMs using synthetic and real data. From the results, the benefits in likelihood and classification accuracy from the proposed model are quantified and analyzed.
The COVID-19 Infodemic had an unprecedented impact on health behaviors and outcomes at a global scale. While many studies have focused on a qualitative and quantitative understanding of misinformation, including sentiment analysis, there is a gap in understanding the emotion-carriers of misinformation and their differences across geographies. In this study, we characterized emotion carriers and their impact on vaccination rates in India and the United States. A manually labelled dataset was created from 2.3 million tweets and collated with three publicly available datasets (CoAID, AntiVax, CMU) to train deep learning models for misinformation classification. Misinformation labelled tweets were further analyzed for behavioral aspects by leveraging Plutchik Transformers to determine the emotion for each tweet. Time series analysis was conducted to study the impact of misinformation on spatial and temporal characteristics. Further, categorical classification was performed using transformer models to assign categories for the misinformation tweets. Word2Vec+BiLSTM was the best model for misinformation classification, with an F1-score of 0.92. The US had the highest proportion of misinformation tweets (58.02%), followed by the UK (10.38%) and India (7.33%). Disgust, anticipation, and anger were associated with an increased prevalence of misinformation tweets. Disgust was the predominant emotion associated with misinformation tweets in the US, while anticipation was the predominant emotion in India. For India, the misinformation rate exhibited a lead relationship with vaccination, while in the US it lagged behind vaccination. Our study deciphered that emotions acted as differential carriers of misinformation across geography and time. These carriers can be monitored to develop strategic interventions for countering misinformation, leading to improved public health.