Time Series Analysis Research Articles

The incremental design of a machine learning framework for medical records processing.

This work presents the development and evaluation of coordn8, a web-based application that streamlines fax processing in outpatient clinics using a "human-in-the-loop" machine learning framework. We demonstrate the effectiveness of the platform at reducing fax processing time and producing accurate machine learning inferences across the tasks of patient identification, document classification, spam classification, and duplicate document detection. We deployed coordn8 in 11 outpatient clinics and conducted a time savings analysis by observing users and measuring fax processing event logs. We used statistical methods to evaluate the machine learning components across different datasets to show generalizability. We conducted a time series analysis to show variations in model performance as new clinics were onboarded and to demonstrate our approach to mitigating model drift. Our observation analysis showed a mean reduction in individual fax processing time by 147.5 s, while our event log analysis of over 7000 faxes reinforced this finding. Document classification produced an accuracy of 81.6%, patient identification produced an accuracy of 83.7%, spam classification produced an accuracy of 98.4%, and duplicate document detection produced a precision of 81.0%. Retraining document classification increased accuracy by 10.2%. coordn8 significantly decreased fax-processing time and produced accurate machine learning inferences. Our human-in-the-loop framework facilitated the collection of high-quality data necessary for model training. Expanding to new clinics correlated with performance decline, which was mitigated through model retraining. Our framework for automating clinical tasks with machine learning offers a template for health systems looking to implement similar technologies.

Streamflow in the United States: Characteristics, trends, regime shifts, and extremes

Long-term streamflow observations contain essential information for understanding hydrological changes and managing water resources. A continental-scale dataset or analysis of temporal streamflow change is still lacking across hydrologic gauges in the Conterminous United States (CONUS). Here, we compiled 70 years of streamflow records from 1951 to 2021 at ~ 8000 hydrologic stations across the CONUS and characterized temporal trends, regime shifts, and extreme events using a Bayesian time series analysis algorithm. We found that the occurrences of sudden streamflow changes (e.g., regime shifts and extreme events) have been increasing with time across the CONUS. In addition, we derived 181 streamflow indicators that are valuable for hydrological and biological applications, such as the duration and frequency of high or low streamflow events. The Mississippi River Basin, especially the middle and lower parts, was a hot spot of high-frequency high-flow events. Overall, we anticipate the dataset generated here offers valuable information for understanding and quantifying changes in water resources across the CONUS.

Time-series analysis of radon concentrations for Bengaluru's atmosphere.

The Fast Fourier Transform (FFT) analysis of the activity concentration of radon and selected meteorological parameters was carried out at Department of Physics, Bangalore University, Bengaluru (12056'44"N, 77030'25″E, 840m above MSL). All of the measured parameters, with the exception of pressure, showed a clear diurnal trend, which can be explained by the presence of typical atmospheric processes such as temperature inversion in the morning and greater vertical mixing in the afternoon. Radon's time series has a latent memory of sub-diurnal cycles, as shown via FFT analysis. The monthly average radon has higher levels of activity during winter months compared with monsoon and summer months. Days during the monsoon season had the lowest radon activity, which may be ascribed to the fact that less radon was being exhaled from the soil as a result of the rain. Radon was recorded at 8.06±0.56Bq/m3, temperature at 28.9°C, humidity at 55.2% and pressure at 918mbar.

Time series analysis of COVID-19's impact on physician and dentist visits in Iran

This study aimed to assess the impact of the COVID-19 pandemic on general practitioner (GP), specialist, and dentist visits among 40 million Iranians covered by the Social Security Organization (SSO). A monthly interrupted time series analysis was conducted over a period of 72 months, including—47 months before the pandemic and 25 months after its onset. The outcomes variables were monthly number of GP, specialist, and dentist visits per 1000 SSO-insured individuals. The analysis was performed by total visits, visits to the SSO direct sector, and visits to the indirect sectors. The study found that in the first month of the pandemic, the number of visits per 1000 insured individuals significantly decreased for visits to GPs (by 51.12, 95% CI −64.42 to −37.88), visits to specialists (by 39.11, 95% CI −51.61 to −26.62), and visits to dentists (by 6.67, 95% CI −8.55 to −4.78). However, during the subsequent months of the pandemic, there was a significant increase in the number of monthly visits for all three categories, with GPs experiencing the highest increase (1.78 visits per 1000 insured), followed by specialists (1.32 visits per 1000 insured), and dentists (0.05 visits per 1000 insured). Furthermore, prior to the pandemic, the number of monthly GP visits per 1000 insured individuals was statistically significantly lower in the indirect sector compared to the direct sector (45.79, 95% CI −52.69 to −38.89). Conversely, the direct sector exhibited lower rates of specialist visits (25.84 visits per 1000 insured individuals, 95% CI 22.87 to 28.82) and dentist visits (0.75 visits per 1000 insured individuals, 95% CI 0.12 to 1.36) compared to the indirect sector. Additionally, the study found that in the first month of the pandemic, the monthly number of GP visits in the indirect sector significantly increased by 34.44 times (95% CI 24.81 to 44.08) compared to the direct sector. For specialist visits and dentist visits, the increase was 3.41 (95% CI −5.87 to 12.69) and 5.01 (95% CI 3.48 to 6.53) per 1000 insured individuals, respectively. Overall, the findings of this study demonstrate statistically significant disruptions in GP, specialist, and dentist visits during the COVID-19 pandemic, although some recovery was observed. Both the direct and indirect sectors experienced decreased visits.

Validation of Citizen Science Meteorological Data: Can They Be Considered a Valid Help in Weather Understanding and Community Engagement?

Citizen science has emerged as a potent approach for environmental monitoring, leveraging the collective efforts of volunteers to gather data at unprecedented scales. Within the framework of the I-CHANGE project, MeteoTracker, a citizen science initiative, was employed to collect meteorological measurements. Through MeteoTracker, volunteers contributed to a comprehensive dataset, enabling insights into local weather patterns and trends. This paper presents the analysis and the results of the validation of such observations against the official Italian civil protection in situ weather network, demonstrating the effectiveness of citizen science in generating valuable environmental data. The work discusses the methodology employed, including data collection and statistical analysis techniques, i.e., time-series analysis, spatial and temporal interpolation, and correlation analysis. The overall analysis highlights the high quality and reliability of citizen-generated data as well as the strengths of the MeteoTracker platform. Furthermore, our findings underscore the potential of citizen science to augment traditional monitoring efforts, inform decision-making processes in environmental research and management, and improve the social awareness about environmental and climate issues.

Complex Network Model of Global Financial Time Series Based on Different Distance Functions

By constructing a complex network model grounded in time series analysis, this study delves into the intricate relationships between the stock markets of 18 countries worldwide. Utilizing 31 distinct time series distance functions to formulate the network, we employ Hamming distance to quantify the resemblance between networks derived from different distance functions. By modulating the network density through distance percentiles (p=0.1, 0.3, 0.5), we demonstrate the similarity of various distance functions across multiple density levels. Our findings reveal that certain distance functions exhibit high degrees of similarity across varying network densities, suggesting their potential for mutual substitution in network construction. Furthermore, the centroid network identified via hierarchical cluster analysis highlights the similarities between the stock markets of different nations, mirroring the intricate interconnections within the global financial landscape. The insights gained from this study offer crucial perspectives for comprehending the intricate network structure of global financial time series data, paving the way for further analysis and prediction of global financial market dynamics.

Spatial and Temporal Evolution of Seasonal Sea Ice Extent of Hudson Strait, Canada, 1971–2018

The temporal and spatial variation in seasonal sea ice in Hudson Strait is examined using time series and spatial clustering analyses. For the period from 1971 to 2018, a time series of sea ice breakup and freeze-up dates and ice-free season length at twenty-four grid points were generated from sea ice charts derived from satellite and other data. These data were analyzed temporally and spatially. The temporal analyses indicated an unambiguous response to a warming climate with statistically significant earlier breakup dates, later freeze-up dates, and longer ice-free seasons, that were statistically linked to coincident regional surface air temperatures. The rate of change in freeze-up dates and ice-free season length was particularly strong in the early 2000s and less so in the 2010s. There was evidence that breakup date behaviour was not only coincident with regional temperatures but likely with temperature and ice conditions of the previous year. Later freeze-up dates were directly linked to earlier breakup dates using detrended time series. Spatial clustering analysis on the Hudson Strait gridded sea ice data revealed distinctive signatures for Ungava Bay, Frobisher Bay, and for grid points close to the shore and a clear linkage to the underlying circulation of Hudson Strait.

Modeling epidemic dynamics using Graph Attention based Spatial Temporal networks.

The COVID-19 pandemic and influenza outbreaks have underscored the critical need for predictive models that can effectively integrate spatial and temporal dynamics to enable accurate epidemic forecasting. Traditional time-series analysis approaches have fallen short in capturing the intricate interplay between these factors. Recent advancements have witnessed the incorporation of graph neural networks and machine learning techniques to bridge this gap, enhancing predictive accuracy and providing novel insights into disease spread mechanisms. Notable endeavors include leveraging human mobility data, employing transfer learning, and integrating advanced models such as Transformers and Graph Convolutional Networks (GCNs) to improve forecasting performance across diverse geographies for both influenza and COVID-19. However, these models often face challenges related to data quality, model transferability, and potential overfitting, highlighting the necessity for more adaptable and robust approaches. This paper introduces the Graph Attention-based Spatial Temporal (GAST) model, which employs graph attention networks (GATs) to overcome these limitations by providing a nuanced understanding of epidemic dynamics through a sophisticated spatio-temporal analysis framework. Our contributions include the development and validation of the GAST model, demonstrating its superior forecasting capabilities for influenza and COVID-19 spread, with a particular focus on short-term, daily predictions. The model's application to both influenza and COVID-19 datasets showcases its versatility and potential to inform public health interventions across a range of infectious diseases.

Impact of the COVID-19 pandemic on antidepressant consumption in the Central region of Portugal: interrupted time series.

To evaluate the impact of the pandemic on the consumption of antidepressive agents in Central Portugal. To estimate the causal effect of the pandemic an interrupted time series analysis was conducted. Data of antidepressant drugs monthly dispensed in community pharmacies between Jan-2010 and Dec-2021 were provided by the regional Health Administration. Anti-Parkinson dopaminergic agents and statins, theoretically not influenced by COVID-19 pandemics, were used as comparator series. The number of packages was converted into defined daily doses and presented as defined daily doses/1000 inhabitants/day. A Bayesian structural time-series model with CausalImpact on R/RStudio was used to predict the counterfactual. Analyses with different geographical granularity (9 sub-regions and 78 municipalities) were performed. When compared to counterfactual, regional consumption non-significantly increased after the pandemic declaration, with a relative effect of + 1.30% [95%CI -1.6%:4.2%]. When increasing the granularity, differences appeared between sub-region with significant increases in Baixo Mondego + 6.5% [1.4%:11.0%], Guarda + 4.4% [1.1%:7.7%] or Cova da Beira + 4.1% [0.17%:8.3%], but non-significant variation in the remaining 6 sub-regions. Differences are more obvious at municipality level, ranging from increases of + 37.00% [32.00%:42.00%] to decreases of -11.00% [-17.00%:-4.20%]. Relative impact positively correlated with percentage of elderly in the municipality (r = 0.301; p = 0.007), and negatively with population density (r=-0.243; p = 0.032). No other predicting variables were found. Antidepressant consumption suffered very slight variations at regional level after the COVID-19 pandemic declaration. Analysis with higher granularity allowed identifying municipalities with higher impact (increase or decrease). The absence of clear association patterns suggests other causal hypotheses of the differences.

Investigating the complex landscape of climate finance in least developed countries (LDCs)

This study aimed to investigate the complex landscape of climate finance, assessing the adequacy, predictability, and implications for sustainable development in least developed countries (LDCs). This study is motivated by the pressing need to assess the adequacy, predictability, and implications of climate finance for sustainable development in least developed countries (LDCs). Employing an econometric framework, this study utilizes ARIMA models to analyze time series data (from 2000 to 2021) on climate finance. The analysis revealed a notable gap between the needed and actual climate funding received by LDCs. Despite an annual requirement of $93.7 billion according to the UK-based International Institute for Environment and Development (IIED), LDCs have only received an average of $14.8 billion annually since 2015. The study suggests that climate funding for LDCs lacks predictability and falls short in meeting their needs, potentially facing an 80% decrease by 2030 under certain scenarios. It advocates for a strategic revamp in climate finance mechanisms to ensure adequacy and predictability, urging policymakers and international funding bodies to adopt more robust, fair, and needs-based approaches to climate financing. This research emphasizes the responsibility of developed nations and global agencies in bridging the considerable funding gap faced by LDCs. By integrating advanced forecasting techniques with a comprehensive analysis of global economic and political factors, this study sheds light on the challenges LDCs encounter in securing stable and sufficient climate finance, stressing the urgency for systemic reforms in global climate finance policies.

S86 - Changes in Opioid and Other Analgesic Prescribing Following Voluntary and Mandatory Prescription Drug Monitoring Program Implementation: A Time Series Analysis

S86 - Changes in Opioid and Other Analgesic Prescribing Following Voluntary and Mandatory Prescription Drug Monitoring Program Implementation: A Time Series Analysis

M147 - Assessing the Impact of Medical Expenditure Supplements Program on the Enrolling and Retention of Patients Attending Methadone Maintenance Treatment Clinics in Taiwan: An Interrupted Time Series Analysis

M147 - Assessing the Impact of Medical Expenditure Supplements Program on the Enrolling and Retention of Patients Attending Methadone Maintenance Treatment Clinics in Taiwan: An Interrupted Time Series Analysis

T58 - Initiations of Safer Supply Hydromorphone Increased During the COVID-19 Pandemic in Ontario: An Interrupted Time-Series Analysis

T58 - Initiations of Safer Supply Hydromorphone Increased During the COVID-19 Pandemic in Ontario: An Interrupted Time-Series Analysis

Different factors affecting the real estate market in the Egyptian context

Purpose Real estate projects are capital-intensive and deeply intertwined with economic factors, making them subject to various influences besides local housing needs. This paper aims to comprehensively understand the dynamics of the Egyptian real estate market, examining real estate cycles, driving factors and their correlation and scale of impact. Design/methodology/approach The study conducts a literature review to explore real estate cycles and their driving factors, along with the relationship between real estate and macroeconomic cycles. It then delves into the dynamics of the Egyptian real estate market, followed by a time series analysis that incorporates five key indicators: economic indicator, demand indicator, supply indicator, capital flow indicator and cost indicator over a 12-year interval (2012–2023), to examine short-term cycle factors, followed by correlation and multi-linear regression analysis to elucidate interrelations among these factors. Findings Through measuring and comparing the prementioned indicators with different economic and social events, the study paints a comprehensive picture of the macroeconomic environment and the real estate cycle in Egypt. Where demand has been found to be more sensitive and directly affected by macroeconomic factors than the supply. With the economic factor as the factor with the highest impact, especially in times of economic fluctuations, the impact has been immediate and short-term. These findings support the idea that the demand in Egypt is speculative, laying a threat of longer recession periods in the long term and having greater and more direct impact. Originality/value This paper contributes to the understanding of the Egyptian real estate market by integrating insights from real estate cycles, macroeconomics and specific market dynamics. The application of time series, correlation and multi-linear regression analysis provides a nuanced understanding of the interplay between several factors shaping the real estate cycle. Ultimately, the findings offer valuable insights for decision-makers involved in urban development planning, facilitating more informed and precise decision-making processes.

Using Bigdata for Choosing the Right Forecasting Method, Dataset and Period in a Time Series Analysis

Nowadays especially production companies gathering a huge data due to their daily transactions on the own systems. Production companies should handle this raw data as handling the raw materials too. Today, scientific studies carried out for this purpose are gathered under the title of BigData. The BigData creates many helps to companies’ competitive advantages according to their competitors. For this view, the purpose of this study was to determine the best demand forecasts method and forecasting period by using BigData at forest production industry. Using the time series analysis module of the WEKA program, the algorithm and data set providing the most accurate estimate for each of the selected decor papers were determined. As a result, it is thought that this study will provide a route map for about choosing right data period and forecasting method for the forest products.

Higher temperatures are associated with increased asthma-related emergency department visits among children: a time series analysis of environmental exposures in Montreal, Canada

Higher temperatures are associated with increased asthma-related emergency department visits among children: a time series analysis of environmental exposures in Montreal, Canada

Analyzing pain patterns in the emergency department: Leveraging clinical text deep learning models for real-world insights

ObjectiveTo determine the incidence of patients presenting in pain to a large Australian inner-city emergency department (ED) using a clinical text deep learning algorithm. Materials and methodsA fine-tuned, domain-specific, transformer-based clinical text deep learning model was used to interpret free-text nursing assessments in the electronic medical records of 235,789 adult presentations to the ED over a three-year period. The model classified presentations according to whether the patient had pain on arrival at the ED. Interrupted time series analysis was used to determine the incidence of pain in patients on arrival over time. We described the changes in the population characteristics and incidence of patients with pain on arrival occurring with the start of the Covid-19 pandemic. Results55.16% (95%CI 54.95%–55.36%) of all patients presenting to this ED had pain on arrival. There were differences in demographics and arrival and departure patterns between patients with and without pain. The Covid-19 pandemic initially precipitated a decrease followed by a sharp, sustained rise in pain on arrival, with concurrent changes to the population arriving in pain and their treatment. DiscussionApplying a clinical text deep learning model has successfully identified the incidence of pain on arrival. It represents an automated, reproducible mechanism to identify pain from routinely collected medical records. The description of this population and their treatment forms the basis of intervention to improve care for patients with pain. The combination of the clinical text deep learning models and interrupted time series analysis has reported on the effects of the Covid-19 pandemic on pain care in the ED, outlining a methodology to assess the impact of significant events or interventions on pain care in the ED. ConclusionApplying a novel deep learning approach to identifying pain guides methodological approaches to evaluating pain care interventions in the ED, giving previously unavailable population-level insights.

Examining geographical inequalities for malaria outcomes and spending on malaria in 40 malaria-endemic countries, 2010–2020

BackgroundWhile substantial gains have been made in the fight against malaria over the past 20 years, malaria morbidity and mortality are marked by inequality. The equitable elimination of malaria within countries will be determined in part by greater spending on malaria interventions, and how those investments are allocated. This study aims to identify potential drivers of malaria outcome inequality and to demonstrate how spending through different mechanisms might lead to greater health equity.MethodsUsing the Gini index, subnational estimates of malaria incidence and mortality rates from 2010 to 2020 were used to quantify the degree of inequality in malaria burden within countries with incidence rates above 5000 cases per 100,000 people in 2020. Estimates of Gini indices represent within-country distributions of disease burden, with high values corresponding to inequitable distributions of malaria burden within a country. Time series analyses were used to quantify associations of malaria inequality with malaria spending, controlling for country socioeconomic and population characteristics.ResultsBetween 2010 and 2020, varying levels of inequality in malaria burden within malaria-endemic countries was found. In 2020, values of the Gini index ranged from 0.06 to 0.73 for incidence, 0.07 to 0.73 for mortality, and 0.00 to 0.36 for case fatality. Greater total malaria spending, spending on health systems strengthening for malaria, healthcare access and quality, and national malaria incidence were associated with reductions in malaria outcomes inequality within countries. In addition, government expenditure on malaria, aggregated government and donor spending on treatment, and maternal educational attainment were also associated with changes in malaria outcome inequality among countries with the greatest malaria burden.ConclusionsThe findings from this study suggest that prioritizing health systems strengthening in malaria spending and malaria spending in general especially from governments will help to reduce inequality of the malaria burden within countries. Given heterogeneity in outcomes in countries currently fighting to control malaria, and the challenges in increasing both domestic and international funding allocated to control and eliminate malaria, the efficient targeting of limited resources is critical to attain global malaria eradication goals.

Gap-filling techniques applied to the GOCI-derived daily sea surface salinity product for the Changjiang diluted water front in the East China Sea

Abstract. The spatial and temporal resolutions of contemporary microwave-based sea surface salinity (SSS) measurements are insufficient. Thus, we developed a gap-free gridded daily SSS product with higher spatial and temporal resolutions, which can provide information on short-term variability in the East China Sea (ECS), such as the front changes by Changjiang diluted water (CDW). Specifically, we conducted gap-filling for daily SSS products based on the Geostationary Ocean Color Imager (GOCI) with a spatial resolution of 1 km (0.01°), using a machine learning approach during the summer seasons from 2015 to 2019. The comparison of the Soil Moisture Active Passive (SMAP), Copernicus Marine Environment Monitoring Service (CMEMS), and Hybrid Coordinate Ocean Model (HYCOM) SSS products with the GOCI-derived SSS over the entire SSS range showed that the SMAP SSS was highly consistent, whereas the HYCOM SSS was the least consistent. In the < 31 psu range, the SMAP SSS was still the most consistent with the GOCI-derived SSS (R2=0.46; root mean squared error: RMSE = 2.41 psu); in the > 31 psu range, the CMEMS and HYCOM SSS products showed similar levels of agreement with that of the SMAP SSS. We trained and tested three machine learning models – the fine trees, boosted trees, and bagged trees models – using the daily GOCI-derived SSS as output, including the three SSS products, environmental variables, and geographical data. We combined the three SSS products to construct input datasets for machine learning. Using the test dataset, the bagged trees model showed the best results (mean R2=0.98 and RMSE = 1.31 psu), and the models that used the SMAP SSS as input had the highest level. For the dataset in the > 31 psu range, all the models exhibited similarly reasonable performances (RMSE = 1.25–1.35 psu). The comparison with in situ SSS data, time series analysis, and the spatial SSS distribution derived from models showed that all the models had proper CDW distributions with reasonable RMSE levels (0.91–1.56 psu). In addition, the CDW front derived from the model gap-free daily SSS product clearly demonstrated the daily oceanic mechanism during the summer season in the ECS at a detailed spatial scale. Notably, the CDW front in the zonal direction, as captured by the Ieodo Ocean Research Station (I-ORS), moved approximately 3.04 km d−1 in 2016, which is very fast compared with the cases in other years. Our model yielded a gap-free gridded daily SSS product with reasonable accuracy and enabled the successful recognition of daily SSS fronts at the 1 km level, which was previously not possible with ocean color data. Such successful application of machine learning models can further provide useful information on the long-term variation of daily SSS in the ECS. The gridded gap-free SSS dataset at 0.01°×0.01° spatial resolution is freely available at https://doi.org/10.22808/DATA-2023-2 (Shin et al., 2023).

Understanding Salinity Intrusion and Residence Times in a Small-Scale Bar-Built Estuary under Drought Scenarios: The Maipo River Estuary, Central Chile

The Maipo River estuary is a low-inflow bar-built estuary that includes a protected wetland, which harbors a rich ecosystem. The estuary and wetland have been threatened by a persistent drought for more than a decade, which has resulted in greater salinity intrusion and increased residence times. Previous studies have described salinity and pollutants in estuaries; however, almost all have focused on deeper and/or wider estuaries with dimensions much larger than those of the small-scale Maipo River estuary. In this study, we used the numerical model FVCOM to simulate the dynamics of the Maipo River estuary under drought scenarios and explored the interactions between river discharge and tides in terms of saline intrusion and particle dispersal. The model was validated against observations collected during a field campaign near the river mouth. The simulations successfully reproduced the water surface elevation but underestimated salinity values, such that the vertical salinity structure observed in the field was not captured by the model in this shallow and morphologically complex estuary. Consequently, our model results provide qualitative insight related to salinity and baroclinic dynamics. Results of maximum saline intrusion showed an exponential decay with increasing river discharge, and the analysis of salinity intrusion time series revealed that droughts may cause permanent non-zero salinity levels in the estuary, potentially affecting ecological cycles. The incorporation of passive tracers showed that decreasing river discharge increases the residence time of particles by allowing the tracers to re-enter the estuary. Model results showed the formation of accumulation zones (hotspots) in the shallower zones of the estuary.