Using a manifold-based approach to extract clinical codes associated with winter respiratory viruses at an emergency department

Every winter, respiratory viruses put most Emergency Departments (ED) around the world under intense pressure. To reduce the consequent stress for hospitals, anticipation of the massive increase of intakes for illness-based symptoms is essential. As the Covid-19 2020 pandemic clearly illustrates, patients are not systematically tested. The ED staff therefore has no real-time knowledge of the presence of the virus in the patients flow. To address this issue, we propose here to use the hospital's laboratory-confirmed database as an attractor for the manifold-based approach for clustering the clinical codes associated with respiratory viruses. We propose a new framework based on the embedding of time series onto the Stiefel manifold, coupled with a density-based clustering algorithm (HDBSCAN) enhanced by a reduction of dimension (UMAP) for the clustering on that manifold. In particular, we show, based on real data sets of two academic hospitals in France, the significant benefits of using geometrical approaches for time series clustering as compared to traditional methods.

Where do people meet? Time-series clustering for social interaction levels in daily-life spaces during the COVID-19 pandemic

The persistence of spatial segregation with respect to income and race is well documented. However, assessment of spatial segregation in daily activities is challenging due to the limited availability of human movement data. With the ubiquitous availability of mobile phones and location-based service applications, human movement data has become widely available. It is now possible to explore spatial interactions and assess the extent of social segregation in daily activity spaces. Using Los Angeles County as our case study, we perform a temporal analysis by conducting K-means time-series clustering using mobile phone data to examine social interaction levels among various sociodemographic groups during the COVID-19 pandemic. Selected sociodemographic variables are assessed among the identified time-series clusters. We find a strong association between sociodemographic characteristics and social interaction levels, potentially leading to disparate exposures to the risk from COVID-19. Socially disadvantaged populations tend to be more segregated from other groups in daily activities, and the COVID-19 pandemic increases the disparity. Low-income and ethnic minority populations became more isolated from Whites and the more economically affluent during the COVID-19 pandemic. Policies that aim to encourage social interactions and mitigate segregation effectively should further consider people's sociodemographic variables and relevant neighborhood characteristics. © 2023 Elsevier Ltd

Modeling the Optimal Grocery Store Trading Area Using Machine Learning Methods

Over the past few years, the COVID-19 pandemic has significantly transformed consumer behavior, which has undoubtedly affected a large number of industries. Food retail was among the sectors where the effect was significant and led to the transformation of the approach to customer interaction. A large part of consumers began to use online delivery services more, and key players were able to provide delivery of products with their own delivery services or third-party on-demand courier service companies. Undoubtedly, in addition to operational changes in retailers' business model, this also affected their investment activities. Some key players began to reduce their trading floor areas to increase financial efficiency and look for options to work in a convenience store format. In our research, we offer an approach for making the right investment decisions when opening a new store to balance financial metrics and customer satisfaction indicators, which is a key sales driver for the segment of customers who substitute delivery service for brick-and-mortar store visits. Using Machine Learning methods, we solve the task of scenario modeling of revenue and operational efficiency metrics for different areas of the store's trading floor, which allows us to identify the optimal choice for the retailer. Using traffic metrics during peak operation hours, we determine the minimal density of the trading area that will not lead to a decrease in the activity of guests inside the store. Such an approach allows us to evaluate the best format of the store, forecast the object's revenue, and recommend investment project parameters. © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

The Analysis of Mobility Patterns during the COVID-19 Pandemic in Thailand Using Time Series Clustering

The COVID-19 pandemic has affected the lives, health, economics, and travel of all nations, including Thailand. The purpose of this study is to investigate human mobility patterns during the pandemic. We opted to use the public transportation data from January 1st, 2020 until September 28th, 2022 collected from the Ministry of Transport, Thailand as a data source. We conducted a time series study on trend and seasonality patterns, as well as clustering analysis. It can be concluded that public buses and Bangkok electric trains, nationwide state trains and domestic air travel are the two pairs of public transportation with the most similar usage patterns. Moreover, the majority of personal car travel patterns are quite similar to public buses and Bangkok electric trains during some periods. © 2023 IEEE.

Unsupervised analysis of COVID-19 pandemic evolution in brazilian states: Vaccination Scenario.

Brazil is one of the countries with the worst response against the pandemic scenario of coronavírus. At the beginning we were on average with 4000 deaths in a 24 hours period. In the course of this situation, large amounts of health and medicine datasets were being generated in real time, requiring effective ways to extract information and discover patterns that can help in the fight against this disease. And even more important is to monitor the progress of prophylactic measures and whether they are being effective in reducing the spread of the virus. Thus, the aim of this study is to analyze how the coronavirus has different ways to evolve in each Brazilian state with the influences of the vaccination process. To achieve this goal, the time series Clustering Technique based on a K-Means variation was applied, with the similarity metric Dynamic Time Warping (DTW). We produced this study using the data reported by the Ministry of Health in Brazil, referring to deaths per 100k inhabitants and all vaccination data available. Our results indicate an unevenly occurring vaccination and the need to identify other associated patterns with human development indices and other socio-economic indicators, being this the first analysis developed in the country, under the goals above.

Alone We Can Do So Little, Together We Cannot Be Detected †

It is no longer possible to imagine our everyday life without time series data. This includes, for example, market developments, COVID-19 cases, electricity prices, and other data from a wide variety of domains. An important task in the analysis of these data is the detection of anomalies. In most cases, this is accomplished by examining individual time series. In our work, we use the techniques of cluster analysis to establish a relationship between time series and groups of time series. This relationship allows us to observe the development of time series in their entirety, thereby gaining additional insights. Our approach identifies outliers with a real-world reference and enables the user to locate outliers without prior knowledge. To underline the strengths of our approach, we compare our method with another known method on two real-world datasets. We found that our solution needs significantly fewer calculations, produces more reasonable results, and can be applied to real-time data. Moreover, our method detected additional outliers, whose occurrence could be explained by real events. © 2022 by the authors.

Air pollution rebound and different recovery modes during the period of easing COVID-19 restrictions.

Although COVID-19 lockdown policies have improved air quality in numerous countries, there is a lack of empirical evidence on the extent to which recovery has resulted in air pollution rebound, and the differences and similarities among regions' recovery modes during the period of easing COVID-19 restrictions. Here, we used daily air quality data and the recovery index constructed by a city-pair inflow index for 119 cities in China to quantify the impact of recovery on air pollution from March 2 to October 30, 2020. Findings show that recovery has significantly increased air pollution. When the recovery level increased by 10 %, the concentration of PM2.5, SO2, and NO2 respectively deteriorated by 1.10, 0.33, 1.25 µg/m3, and the average growth rates of three air pollutants were about 3 %-6 %. Moreover, we used the counterfactual framework and time series clustering with wavelet transform to cluster the rebound trajectory of air pollution for 17 provinces into five recovery modes. Results show that COVID-19 has further intensified regional differentiations in economic development ability and green recovery trend. Three northwestern provinces dependent on their resource endowments belong to energy-intensive recovery mode, which have experienced a sharp rebound of air pollution for two months, thereby making green recovery more challenging to achieve. Three regions with a diversified industrial structure are in industrial-restructuring recovery mode, which has effectively returned to a normal level through adjusting industrial structure and technological innovation. Owing to local policies and the outbreak of COVID-19 in other countries, six provinces in policy-oriented and international trade-oriented recovery modes have not fully recovered to the level without COVID-19 until October 2020. The result highlights the importance of diversifying industrial structure, technological innovation, policy flexibility and industrial upgrading for different recovery modes to achieve long-term green recovery in the future.

Statistical methods for policy-relevant questions in health and criminology

The logistics of policy implementation can lead to a delay from when the actual change in behavior occurs, leading to a shift in a time series. Using change point analysis allows for the data to determine where a change in mean, or other parameters, occurred. But when policy is implemented across multiple locations, how can a researcher understand where change points are occurring at across all locations? Can those locations be grouped together based on their change point? We propose a methodology for clustering panels of nonlinear time series and develop diagnostics to assess the clustering. The change point component of the methodology allows for trends and point anomalies to be detected for each time series. This methodology incorporates spatial and demographic information from the locations into the clustering aspect of the methodology. In a practical application of our methodology, we investigate when average counts of emergency department (ED) visits change related to when the Affordable Care Act was enacted, using monthly time series from 88 locations. Using the diagnostic measures developed and innovative data analysis techniques we understand the groupings of these locations and where in time these groups were changing. In another data application we investigated the impact COVID-19 had on crime rates in the city of Chicago. Using our methodology and data visualization tools, we examined if neighborhoods experience a reduction in crime through their change points and how to group these time series together.This paper also explores the use of Gaussian graphical models to understand metabolic networks to assist in the development of new targeted assays. A metabolite can be measured through a well-developed panel, called a targeted assay, or through a mass spectrometer reading. The mass spectrometer measure, an untargeted panel, is poorly measured but can detect all metabolites present in the sample unlike the targeted panel which only measures these few well-studied metabolites. Given the high cost of targeting a metabolite, it is important to investigate the benefits of a possible addition of a metabolite to a targeted panel. We developed a model based on the determination of successful targeting of a metabolite using variables related to the metabolite in the network. (PsycInfo Database Record (c) 2022 APA, all rights reserved)

The Clustering Structure of the COVID-19 Outbreak in Global Scale

Spreading of novel coronavirus disease started in China and moved to Korea and Japan, then several countries in Europe, and the last step to the countries in the North and South American continents. Since the virus spread worldwide, we simultaneously use all available daily confirmed cases, recovered cases, and death data to cluster countries in time and spatial dimensions after adjusting for population. For this aim, time-series clustering with the dynamic time warping method is implemented and relevant clusters are marked on the world maps for a better visual understanding in this paper. Grouping countries will give an idea of the spread of the virus, guide decision-makers to implement future prevention vaccination policies, and help them generate global solutions against new virus variants. One of the main results obtained from the cluster analysis is that the European, North and South American continents have homogeneous structures regarding the number of daily confirmed cases per million and relatively more heterogeneous regarding the daily number of recoveries per million such that the overwhelming majority of countries are in the very high cluster. The absence of countries from the low or middle clusters indicates that these continents have to fight the virus more fiercely. African and Asian continents are heterogeneous in all cases. Therefore, these continents should focus on country-specific protections to fight against the virus.

How did COVID-19 change what people buy: Evidence from a supermarket chain

This research takes a retrospective view of the COVID-19 pandemic and attempts to accurately measure its impact on sales of different product categories in grocery retail. In total 150 product categories were analyzed using the data of a major supermarket chain in the Netherlands. We propose to measure the pandemic impact by excess sales – the difference of actual and expected sales. We show that the pandemic impact is twofold: (1) There was a large but brief growth at 30.6% in excess sales associated with panic buying across most product categories within a two-week period;and (2) People spending most of their time at home due to imposed restrictions resulted in an estimated 5.4% increase in total sales lasting as long as the restrictions were active. The pandemic impact on different product categories varies in magnitudes and timing. Using time series clustering, we identified eight clusters of categories with similar pandemic impacts. Using clustering results, we project that product categories used for cooking, baking or meal preparation in general will have elevated sales even after the pandemic.

Autoregressive mixture models for clustering time series

Clustering time series into similar groups can improve models by combining information across like time series. While there is a well developed body of literature for clustering of time series, these approaches tend to generate clusters independently of model training, which can lead to poor model fit. We propose a novel distributed approach that simultaneously clusters and fits autoregression models for groups of similar individuals. We apply a Wishart mixture model so as to cluster individuals while modelling the corresponding autocovariance matrices at the same time. The fitted Wishart scale matrices map to cluster-level autoregressive coefficients through the Yule-Walker equations, fitting robust parsimonious autoregressive mixture models. This approach is able to discern differences in underlying autocorrelation variation of time series in settings with large heterogeneous datasets. We prove consistency of our cluster membership estimator, asymptotic distributions of coefficients and compare our approach against competing methods through simulation as well as by fitting a COVID-19 forecast model.

Health care visits during the COVID-19 pandemic: A spatial and temporal analysis of mobile device data.

Transportation disruptions caused by COVID-19 have exacerbated difficulties in health care delivery and access, which may lead to changes in health care use. This study uses mobile device data from SafeGraph to explore the temporal patterns of visits to health care points of interest (POIs) during 2020 and examines how these patterns are associated with socio-demographic and spatial characteristics at the Census Block Group level in North Carolina. Specifically, using the K-medoid time-series clustering method, we identify three distinct types of temporal patterns of visits to health care facilities. Furthermore, by estimating multinomial logit models, we find that Census Block Groups with higher percentages of elderly persons, minorities, low-income individuals, and people without vehicle access are areas most at-risk for decreased health care access during the pandemic and exhibit lower health care access prior to the pandemic. The results suggest that the ability to conduct in-person medical visits during the pandemic has been unequally distributed, which highlights the importance of tailoring policy strategies for specific socio-demographic groups to ensure equitable health care access and delivery.

Data-driven identification of temporal glucose patterns in a large cohort of nondiabetic patients with COVID-19 using time-series clustering.

OBJECTIVE: Hyperglycemia has emerged as an important clinical manifestation of coronavirus disease 2019 (COVID-19) in diabetic and nondiabetic patients. Whether these glycemic changes are specific to a subgroup of patients and persist following COVID-19 resolution remains to be elucidated. This work aimed to characterize longitudinal random blood glucose in a large cohort of nondiabetic patients diagnosed with COVID-19. MATERIALS AND METHODS: De-identified electronic medical records of 7502 patients diagnosed with COVID-19 without prior diagnosis of diabetes between January 1, 2020, and November 18, 2020, were accessed through the TriNetX Research Network. Glucose measurements, diagnostic codes, medication codes, laboratory values, vital signs, and demographics were extracted before, during, and after COVID-19 diagnosis. Unsupervised time-series clustering algorithms were trained to identify distinct clusters of glucose trajectories. Cluster associations were tested for demographic variables, COVID-19 severity, glucose-altering medications, glucose values, and new-onset diabetes diagnoses. RESULTS: Time-series clustering identified a low-complexity model with 3 clusters and a high-complexity model with 19 clusters as the best-performing models. In both models, cluster membership differed significantly by death status, COVID-19 severity, and glucose levels. Clusters membership in the 19 cluster model also differed significantly by age, sex, and new-onset diabetes mellitus. DISCUSSION AND CONCLUSION: This work identified distinct longitudinal blood glucose changes associated with subclinical glucose dysfunction in the low-complexity model and increased new-onset diabetes incidence in the high-complexity model. Together, these findings highlight the utility of data-driven techniques to elucidate longitudinal glycemic dysfunction in patients with COVID-19 and provide clinical evidence for further evaluation of the role of COVID-19 in diabetes pathogenesis.

Spatial robust fuzzy clustering of COVID 19 time series based on B-splines.

The aim of the work is to identify a clustering structure for the 20 Italian regions according to the main variables related to COVID-19 pandemic. Data are observed over time, spanning from the last week of February 2020 to the first week of February 2021. Dealing with geographical units observed at several time occasions, the proposed fuzzy clustering model embedded both space and time information. Properly, an Exponential distance-based Fuzzy Partitioning Around Medoids algorithm with spatial penalty term has been proposed to classify the spline representation of the time trajectories. The results show that the heterogeneity among regions along with the spatial contiguity is essential to understand the spread of the pandemic and to design effective policies to mitigate the effects.