Patterns In Spatiotemporal Data Research Articles

Dynamic multi-fusion spatio-temporal graph neural network for multivariate time series forecasting

Multivariate time series data with multiple points is a kind of spatio-temporal data, and in recent years, spatio-temporal graph neural networks (STGNNs) have become one of the most effective methods for spatio-temporal data prediction. By leveraging graph neural networks and sequence models, STGNNs simultaneously capture the hidden temporal and spatial patterns in spatio-temporal data. However, there exist two key limitations: (1) Most existing STGNNs whether based on static graphs or dynamic graphs, fail to introduce time periodicity into graph structure learning and ignore the importance of multi-relational graph structure learning. (2) Most multivariate time series prediction models with multiple points ignore the relation between multiple attributes. To this end, we propose a novel framework called the Dynamic Multi-fusion Spatio-temporal Graph Neural Network (DMF-STNet), where multi-fusion embodies two meanings: spatio-temporal fusion and multi-attribute fusion, as well as multi-graph fusion and multi-layer fusion. The dynamic periodic graph is proposed to mine time periodicity that is hidden in graph structures, and we further offer multi-relational graph structure learning methods to capture static and dynamic relations. Moreover, we introduce a dual-stage feature aggregation method to aggregate hidden features from both primary and auxiliary attributes. Extensive experiments on three real-world datasets demonstrate the efficacy of our proposed DMF-STNet, and the visualization study proves the interpretability of learned weights.

Assessing Refugees’ Integration via Spatio-Temporal Similarities of Mobility and Calling Behaviors

In Turkey the increasing tension, due to the presence of 3.4 million Syrian refugees, demands the formulation of effective integration policies. Moreover, their design requires tools aimed at understanding the integration of refugees despite the complexity of this phenomenon. In this work, we propose a set of metrics aimed at providing insights and assessing the integration of Syrians refugees, by analyzing a real-world Call Details Records (CDRs) dataset including calls from refugees and locals in Turkey throughout 2017. Specifically, we exploit the similarity between refugees' and locals' spatial and temporal behaviors, in terms of communication and mobility in order to assess integration dynamics. Together with the already known methods for data analysis, we use a novel computational approach to analyze spatiotemporal patterns: Computational Stigmergy, a bio-inspired scalar and temporal aggregation of samples. Computational Stigmergy associates each sample to a virtual pheromone deposit (mark). Marks in spatiotemporal proximity are aggregated into functional structures called trails, which summarize the spatiotemporal patterns in data and allows computing the similarity between different patterns. According to our results, collective mobility and behavioral similarity with locals have great potential as measures of integration, since they are: (i) correlated with the amount of interaction with locals; (ii) an effective proxy for refugee's economic capacity, thus refugee's potential employment; and (iii) able to capture events that may disrupt the integration phenomena, such as social tensions.

Machine Learning for Drug Overdose Surveillance

ABSTRACTWe describe two recently proposed machine learning approaches for discovering emerging trends in fatal accidental drug overdoses. The Gaussian Process Subset Scan (Herlands, McFowland, Wilson, & Neill, 2017) enables early detection of emerging patterns in spatio-temporal data, accounting for both the complex, correlated nature of the data and the fact that detecting subtle patterns requires integration of information across multiple spatial areas and multiple time steps. We apply this approach to 17 years of county-aggregated data for monthly opioid overdose deaths in the New York City metropolitan area, showing clear advantages in the utility of discovered patterns as compared to typical anomaly detection approaches. To detect and characterize emerging overdose patterns that differentially affect a subpopulation of the data, including geographic, demographic, and behavioral patterns (e.g., which combinations of drugs are involved), we apply the Multidimensional Tensor Scan (Neill, 2017) to 8 years of case-level overdose data from Allegheny County, Pennsylvania. We discover previously unidentified overdose patterns which reveal unusual demographic clusters, show impacts of drug legislation, and demonstrate potential for early detection and targeted intervention. These approaches to early detection of overdose patterns can inform prevention and response efforts, as well as understanding the effects of policy changes.

CLUSTERING-BASED APPROACHES TO THE EXPLORATION OF SPATIO-TEMPORAL DATA

Abstract. As one spatio-temporal data mining task, clustering helps the exploration of patterns in the data by grouping similar elements together. However, previous studies on spatial or temporal clustering are incapable of analysing complex patterns in spatio-temporal data. For instance, concurrent spatio-temporal patterns in 2D or 3D datasets. In this study we present two clustering algorithms for complex pattern analysis: (1) the Bregman block average co-clustering algorithm with I-divergence (BBAC_I) which enables the concurrent analysis of spatio-temporal patterns in 2D data matrix, and (2) the Bregman cube average tri-clustering algorithm with I-divergence (BCAT_I) which enables the complete partitional analysis in 3D data cube. Here the use of the two clustering algorithms is illustrated by Dutch daily average temperature dataset from 28 weather stations from 1992 to 2011. For BBAC_I, it is applied to the averaged yearly dataset to identify station-year co-clusters which contain similar temperatures along stations and years, thus revealing patterns along both spatial and temporal dimensions. For BCAT_I, it is applied to the temperature dataset organized in a data cube with one spatial (stations) and two nested temporal dimensions (years and days). By partitioning the whole dataset into clusters of stations and years with similar within-year temperature similarity, BCAT_I explores the spatio-temporal patterns of intra-annual variability in the daily temperature dataset. As such, both BBAC_I and BCAT_I algorithms, combined with suitable geovisualization techniques, allow the exploration of complex spatial and temporal patterns, which contributes to a better understanding of complex patterns in spatio-temporal data.

Triclustering Georeferenced Time Series for Analyzing Patterns of Intra-Annual Variability in Temperature

Clustering is often used to explore patterns in georeferenced time series (GTS). Most clustering studies, however, only analyze GTS from one or two dimension(s) and are not capable of the simultaneous analysis of the data from three dimensions: spatial, temporal, and any third (e.g., attribute) dimension. Here we develop a novel clustering algorithm called the Bregman cuboid average triclustering algorithm with I-divergence (BCAT_I), which enables the complete partitional analysis of 3D GTS. BCAT_I simultaneously groups the data along its dimensions to form regular triclusters. These triclusters are subsequently refined using k means to fully capture spatiotemporal patterns in the data. By applying BCAT_I to time series of daily average temperature in The Netherlands (twenty-eight weather stations from 1992 to 2011), we identified the refined triclusters with similar temperature values along the spatial dimension (weather stations that represent locations) and two nested temporal dimensions (year and day). Geovisualization techniques were then used to display the patterns of intra-annual variability in temperature. Our results show that in the last two thirds of the study period, there is an intense variability of spring and winter temperatures in the northeast and center of The Netherlands. For the same period, an intense variability of spring temperatures is also visible in the southeast of the country. Our results also show that summer temperatures are homogenous across the country for most of the study period. This particular application demonstrates that BCAT_I enables a complete analysis of 3D GTS and, as such, it contributes to a better understanding of complex patterns in spatiotemporal data.

The influence of domain expertise on visual overviews of spatiotemporal data

ABSTRACTOverviews of spatiotemporal data are acknowledged to play an important role in visualization in initiating and supporting geovisualization and exploratory data analysis (EDA). However, relatively little research has focused on the visual overviews themselves, and their potential impacts on EDA outcomes. In a user study, we evaluated the influence of different levels of domain knowledge on the usefulness of four distinct types of static visual overview of spatiotemporal data. Beyond simply orienting users, our results indicate that visual overviews can be important in gaining insights into a data set, for example, in learning about metadata. Although subjects without domain knowledge struggled to judge the quality of their findings, they were as successful at identifying interesting patterns in the data as those with domain expertise. Our results suggest that detailed background knowledge of a data set can actively hinder EDA. Being already familiar with their own data sets, our results highlight the tendency of data experts to disregard findings that do not match their pre-existing domain knowledge. Based on these findings, our conclusions identify a range of potential avenues for future work, including the use of visual overviews that deliberately do not, from first view, reveal the context of the data they show. This later approach could help in cases where domain experts need to see their data with ‘fresh eyes’, and detect interesting patterns in spatiotemporal data before relating the findings to specific knowledge about the data sets and the domain.

Exploration of the use of Bayesian modeling of gradients for censored spatiotemporal data from the Deepwater Horizon oil spill

This paper develops a hierarchical framework for identifying spatiotemporal patterns in data with a high degree of censoring using the gradient process. To do this, we impute censored values using a sampling-based inverse CDF method within our Markov chain Monte Carlo algorithm, thereby avoiding burdensome integration and facilitating efficient estimation of other model parameters. We illustrate use of our methodology using a simulated data example, and uncover the danger of simply substituting a space- and time-constant function of the level of detection for all missing values. We then fit our model to area measurement data of volatile organic compound (VOC) air concentrations collected on vessels supporting the response and clean-up efforts of the Deepwater Horizon oil release that occurred starting April 20, 2010. These data contained a high percentage of observations below the detectable limits of the measuring instrument. Despite this, we were still able to make some interesting discoveries, including elevated levels of VOC near the site of the oil well on June 26th. Using the results from this preliminary analysis, we hope to inform future research on the Deepwater Horizon study, including the use of gradient methods for assigning workers to exposure categories.

Hierarchical spatiotemporal feature extraction using recurrent online clustering

Deep machine learning offers a comprehensive framework for extracting meaningful features from complex observations in an unsupervised manner. The majority of deep learning architectures described in the literature primarily focus on extracting spatial features. However, in real-world settings, capturing temporal dependencies in observations is critical for accurate inference. This paper introduces an enhancement to DeSTIN – a compositional deep learning architecture in which each layer consists of multiple instantiations of a common node – that learns to represent spatiotemporal patterns in data based on a novel recurrent clustering algorithm. Contrary to mainstream deep architectures, such as deep belief networks where layer-by-layer training is assumed, each of the nodes in the proposed architecture is trained independently and in parallel. Moreover, top-down and bottom-up information flows facilitate rich feature formation. A semi-supervised setting is demonstrated achieving state-of-the-art results on the MNIST classification benchmarks. A GPU implementation is discussed further accentuating the scalability properties of the proposed framework.

Mining Volunteered Geographic Information datasets with heterogeneous spatial reference

When the information created online by users has a spatial reference, it is known as Volunteered Geographic Information (VGI). The increased availability of spatiotemporal data collected from satellite imagery and other remote sensors provides opportunities for enhanced analysis of Spatiotemporal Patterns. This area can be defined as efficiently discovering interesting patterns from large data sets. The discovery of hidden periodic patterns in spatiotemporal data could provide unveiling important information to the data analyst. In many applications that track and analyze spatiotemporal data, movements obey periodic patterns; the objects follow the same routes (approximately) over regular time intervals. However, these methods cannot directly be applied to a spatiotemporal sequence because of the fuzziness of spatial locations in the sequence. In this paper, we define the problem of mining VGI datasets with our already established bottom up algorithm for spatiotemporal data.

Discovery of Periodic Patterns in Spatiotemporal Sequences

In many applications that track and analyze spatiotemporal data, movements obey periodic patterns; the objects follow the same routes (approximately) over regular time intervals. For example, people wake up at the same time and follow more or less the same route to their work everyday. The discovery of hidden periodic patterns in spatiotemporal data could unveil important information to the data analyst. Existing approaches for discovering periodic patterns focus on symbol sequences. However, these methods cannot directly be applied to a spatiotemporal sequence because of the fuzziness of spatial locations in the sequence. In this paper, we define the problem of mining periodic patterns in spatiotemporal data and propose an effective and efficient algorithm for retrieving maximal periodic patterns. In addition, we study two interesting variants of the problem. The first is the retrieval of periodic patterns that are frequent only during a continuous subinterval of the whole history. The second problem is the discovery of periodic patterns, whose instances may be shifted or distorted. We demonstrate how our mining technique can be adapted for these variants. Finally, we present a comprehensive experimental evaluation, where we show the effectiveness and efficiency of the proposed techniques