Time Series Analysis Research Articles

Slidingan analysis of analytical signal of non-contact photoplethysmography for assessing heart rate

The paper proposes a method for studying the variability of the subject’s heart rate based on the intellectual analysis of the pulse wave measured with remote photoplethysmography. The logically related stages of the formation of quadrature components based on the Hilbert transform of biomedical signals’ dynamics are presented. Within the framework of modern methods of intellectual analysis of non-stationary time series, realizations of adaptive estimates of instantaneous frequencies and periods of the heartbeat basic tone are obtained.

“The 21st century professionalization?”: online education as an instrument for bolstering individual welfare and societal equality

Abstract Since the start of the last decade of the 20th century, the application of digital technologies with the potential of making substantial contributions to a community’s welfare have led to the creation of the literature of Public Interest Technologies and reinvigorated many areas of focus in the social sciences. A relevant subfield of this literature has been the application of new technologies to education to increase social public benefits. The traditional literature on higher education seems to agree that the expansion of this good throughout a society is desirable because it generates direct and indirect individual gains, as well as aggregate level social benefits in the form of positive externalities. However, some authors have argued that the acquisition of postsecondary education has its own drawbacks: unwanted intellectual influence of instructors on students, barriers to access for students of historically vulnerable sociodemographic groups, as well as students’ perception of a reduced pay-off from going to college. While there have been some indices that online higher education could abate many of these issues, there has been relatively little formal research to test this tool’s impact. Using a quantitative approach, via cross-section and time-series data analysis, the text finds some evidence that online higher education could solve many of higher education’s drawbacks, while also being a feasible approach given the current technological environment of the United States. The text concludes outlining future research using a mixed-methods approach that could be highly valuable to acquire a more comprehensive view and robust evidence of online higher education’s potential of expanding social public benefits.

Multiscale Irreversibility Analysis of Time Series Based on Permutation Jensen–Shannon Distance

In this study, we propose a multiscale permutation Jensen–Shannon distance (MPJSD) to measure irreversibility of complex time series. The new quantifier is based on symbolic permutation pattern and Jensen–Shannon distance. As an alternative, the new method offers the best characterization of the underlying irreversibility on different scales. The ARFIMA process and three dissipative chaotic systems are used to verify the effectiveness of the new method. The numerical results indicate that the MPJSD can unveil subtle and interesting findings on different scales and the permutation Jensen–Shannon distance (PJSD) is scale-dependent. Furthermore, we apply the approach to detect the multiscale irreversibility of ECG and financial data. The underlying irreversible nature of the investigated series is well discriminated. The method here introduced gives a new way to distinguish different degrees of irreversibility.

A "Wise" Intervention to Increase Hand Hygiene Compliance of Nurses in Acute Care Units in US Hospitals: A Multiple Baseline Interrupted Time-Series Evaluation.

Objective: This study tested a 'wise' intervention (quick prompt of a specific psychological mec) in acute care hospital units to improve nurses' hand hygiene compliance (HHC). Design: A multiple baseline design in two medical-surgical teaching hospitals in the United States. Measurements: Hand hygiene data was collected using an electronic compliance monitoring system with sensors placed in doorways and on corresponding soap and alcohol-based hand rub dispensers. The outcome measure was the proportion of opportunities in which HH was undertaken by staff per week in each unit. Intervention: A quick-and-easy psychological prime to reinvigorate professional identity. Methods: Interrupted time series analysis using a quasi-Poisson regression model with statistical process control charts for each unit. Results: A statistically significant increase in HHC rates that was sustained for months post-intervention. However, the patterns by unit were not statistically significant once temporal trends were considered. Other factors, such as the unit type and the use of incentives could have impacted the results. Conclusions: These analyses suggest that the aggregate impact should not be taken as evidence of intervention effectiveness. This study therefore cannot be considered to have provided a strong foundation for use of a 'wise' intervention, despite its relatively small financial, logistical and psychological cost.

Abstract 4136550: In-hospital cardiac arrest can be detected by Innovative smart healthcare syste

Background: Predicting in-hospital cardiac arrest (IHCA) is crucial for preventing patient fatalities. However, it remains challenging due to the irregularities in time series data across various features. Previous studies relied on traditional medical scoring systems, but their accuracy was limited as the models were overly simplistic. Recent research using machine learning methods has shown improved results. In our work, we employ deep learning and time series approaches, which are considered the most powerful models for addressing complex problems in recent years. Method/Research Design: The cohort study took place in a hospital in Taiwan from January 2016 to August 2022, involving 252,000 patients in general wards. Input data encompassed basic information, vital signs, Glasgow Coma Scale scores, medication, etc., to predict IHCA occurrences within 24 hours. Given the immediacy of IHCA events, we utilized left alignment to handle time series. Evaluation metrics included the area under the receiver operating characteristic curve (AUROC). Results: We compared three types of settings. The first was based on scoring systems, where the national early warning system (NEWS) achieved an AUROC of 0.756, and the modified early warning score (MEWS) reached 0.653. In the next setting, based on the data summary within a given time interval, the machine learning models XGBoost and deep learning model TabNet achieved AUROCs of 0.863 and 0.839, respectively. This indicates that machine learning outperforms traditional medical methods. The final setting involved time series analysis, where XGBoost and the Gate Recurrent Unit (GRU) network achieved 0.848 and 0.904, respectively, surpassing all previous settings. Conclusion: Our study demonstrates that learning-based methods outperform rule-based methods in predicting IHCA events. Machine learning algorithms offer more possibilities and have the potential to address complex problems effectively. Moreover, considering time series data rather than summarizing within fixed intervals preserves information and leads to better scores. Leveraging the time series deep learning method, GRU shows promising potential in significantly reducing the risk of IHCA events.

Foreign Finance, FDI-Climate Interaction and Economic Development in Ethiopia: A Time Series Analysis

Developing countries such as Ethiopia, over the years particularly after the cold war, starting from early 1990s have adopted neoliberal policies through reforms to attract foreign finance for development needs in view of inadequacy of domestic capital for public and private investment. However, despite the inflow of foreign capital into the economy, it is still grappling with major development challenges. It is in view of this that this study examines the impact of foreign finance on the economy, and whether FDI-climate interaction limits the potential impact of foreign direct investment (FDI) as an important component of foreign capital, on economic development in Ethiopia. The study further reveals the implications of post-cold war adoption of market-oriented policies in Ethiopia. The study uses autoregressive distributed lag (ARDL) model with data spanning from 1981 to 2020. It is found that FDI has positive and significant impact on economic development in the long run, but the positive effect is revealed to be insignificant in the short run. On the other hand, foreign debt is revealed to have long term negative though insignificant impact on economic development of the country. Furthermore, it is revealed that FDI-climate interaction has negative implication on the economy in both the long run and short run. Therefore, the effectiveness and efficiency of foreign finance on growth and development outcome in Ethiopia is conditional on whether it produces long run sustainable environmental outcome. Also, the adoption of neoliberal policies has not resulted in development of Ethiopia. Hence, the study recommends among others, the need for consideration of absorptive capacity of the country as a recipient economy and in its reforms, as well as the need for the government and development partners to ensure assessment of the long run potential environmental impact of foreign capital before they are deployed on the economy.

Econometrics of Climate Change Research

The econometrics of climate change explores the intricate interplay between economic activity and climate dynamics, offering empirical insights into the causes, impacts, and mitigation of global climate change. This research applies advanced econometric techniques to quantify the economic consequences of environmental changes and evaluate the effectiveness of climate policies. By leveraging panel data models, time-series analysis, and causal inference methods, this study investigates critical topics such as the economic costs of extreme weather events, the efficacy of carbon pricing mechanisms, and the role of renewable energy adoption in mitigating climate risks. Additionally, this research incorporates machine learning techniques to enhance predictive accuracy in modeling future climate-economic interactions. Through rigorous empirical analysis, the study provides actionable recommendations for policymakers to design sustainable strategies that balance economic growth with climate resilience. The findings contribute to a deeper understanding of the socioeconomic dimensions of climate change, emphasizing the importance of evidence-based decision-making in addressing one of humanity’s most pressing challenges.

Nonlinear time series analysis of state-wise COVID-19 in Malaysia using wavelet and persistent homology.

The nonlinear progression of COVID-19 positive cases, their fluctuations, the correlations in amplitudes and phases across different regions, along with seasonality or periodicity, pose challenges to thoroughly examining the data for revealing similarities or detecting anomalous trajectories. To address this, we conducted a nonlinear time series analysis combining wavelet and persistent homology to detect the qualitative properties underlying COVID-19 daily infection numbers at the state level from the pandemic's onset to June 2024 in Malaysia. The first phase involved investigating the evolution of daily confirmed cases by state in the time-frequency domain using wavelets. Subsequently, a topological feature-based time series clustering is performed by reconstructing a higher-dimensional phase space through a delay embedding method. Our findings reveal a prominent 7-day periodicity in case numbers from mid-2021 to the end of 2022. The state-wise daily cases are moderately correlated in both amplitudes and phases during the Delta and Omicron waves. Biweekly averaged data significantly enhances the detection of topological loops associated with these waves. Selangor demonstrates unique case trajectories, while Pahang shows the highest similarity with other states. This methodological framework provides a more detailed understanding of epidemiological time series data, offering valuable insights for preparing for future public health crises.

Comparative Study of Time Series Analysis Algorithms Suitable for Short-Term Forecasting in Implementing Demand Response Based on AMI

This paper compares four time series forecasting algorithms—ARIMA, SARIMA, LSTM, and SVM—suitable for short-term load forecasting using Advanced Metering Infrastructure (AMI) data. The primary focus is on evaluating the applicability and performance of these forecasting models in predicting electricity consumption patterns, which is a critical component for implementing effective demand response (DR) strategies. The study provides a comprehensive analysis of the predictive accuracy, computational efficiency, and scalability of each algorithm using a dataset of real-time electricity consumption collected from AMI systems over a designated period. Through extensive experiments, we demonstrate that each algorithm has distinct strengths and weaknesses depending on the characteristics of the dataset. Specifically, SVM exhibited superior performance in handling nonlinear patterns and high volatility, while SARIMA effectively captured seasonal trends. LSTM showed potential in modeling complex temporal dependencies but was sensitive to hyperparameter settings and required a substantial amount of training data. This research offers practical guidelines for selecting the optimal forecasting model based on data characteristics and application needs, contributing to the development of more efficient and dynamic energy management strategies. The findings highlight the importance of integrating advanced forecasting techniques into smart grid systems to enhance the reliability and responsiveness of DR programs. This study lays a solid foundation for future research on integrating these forecasting models into real-world AMI applications to support effective demand response and grid stability.

Do clean fuel and trade-openness reduce environmental degradation in China: evidence from asymmetric model

IntroductionThis study investigates the relationship between economic growth, clean fuel utilization, trade, and environmental pollution. By focusing on the potential of clean fuels and trade in reducing pollution, this research aims to understand their role in promoting environmental sustainability in China from 1990 to 2020.MethodsThe study employs the Nonlinear Autoregressive Distributed Lags (NARDL) approach to analyze time series data. This method enables the decomposition of impacts from clean fuels and trade on pollution levels, facilitating an understanding of their distinct contributions to environmental outcomes.ResultsFindings indicate that economic growth is associated with an increase in pollution levels, while clean fuel usage exhibits a modest but significant reduction effect. Nonetheless, the relatively small coefficients highlight the need for a greater share of clean fuels in China’s energy mix. Additionally, trade shows a significant negative relationship with pollution emissions, with data reflecting a gradual decline in pollution over time.DiscussionThese results underscore the benefits of clean fuel adoption and trade promotion as effective measures for mitigating pollution. However, they also highlight the necessity for stronger regulatory policies, public awareness initiatives, and investments in sustainable technologies to ensure China’s long-term environmental sustainability.

A novel tucker decomposition driven taxi travel demand forecasting algorithm

<p>This research aims to enhance taxi travel demand forecasting for sustainable urban traffic management and planning. We extend the Seasonal Autoregressive Integrated Moving Average model into a high-dimensional tensor form by treating the urban transport network as a Euclidean space and introducing Tucker decomposition. This novel approach, both theoretically significant and practically applicable, represents time series data as tensors to better capture multimodal structures and correlations, improving predictive accuracy. Tucker decomposition reduces computational complexity and memory requirements, making it ideal for large-scale urban traffic network prediction. Experiments on six real-world datasets show that the model’s MAE and RMSE are reduced by about 39.43% and 27.01% on average, respectively, compared to the baseline model. Notably, the model is computationally very efficient and takes only a relatively short time to train., suitable for real-time traffic management, congestion mitigation, and resource optimization. In summary, this work innovates time series analysis, providing an efficient and precise tool for urban traffic management and planning, contributing to sustainable urban transportation advancement.<strong></strong></p>

Evolution and Research Trends in Physical Culture. A Longitudinal Analysis of Student Diploma Theses from the Didactic Team of Physical Culture Sciences, formerly the Department of Physical Culture, Faculty of Earth Sciences and Spatial Management, 2017-2024

Objective: The aim of this study was to conduct a comprehensive analysis of diploma theses produced in the Department of Physical Culture at the Faculty of Earth Sciences and Spatial Management between 2017-2024, to identify main research trends, thematic evolution, and assess the impact of external factors on research directions in the field of physical culture. Material and Methods: The analysis included 95 diploma theses (82 bachelor's and 13 master's). Mixed research methods were applied, combining quantitative and qualitative analysis. Content analysis, statistical analysis (including chi-square tests, analysis of variance, linear regression), and advanced data analysis methods (LDA topic modeling, social network analysis, SARIMA time series analysis) were conducted. Additionally, measures of thematic diversity and supervisor specialization indicators were applied. Results: A statistically significant upward trend in the number of theses (β = 1.1875, p = 0.0134) and significant seasonality of defenses were found. Five main thematic areas were identified: physical activity and health (29.5%), tourism (23.2%), sport (21.1%), health (15.8%), and recreation (10.5%). LDA analysis revealed more detailed categories, indicating growing interdisciplinarity of research. A significant impact of the COVID-19 pandemic on thesis topics was observed in 2020-2021. Social network analysis showed a key role of certain supervisors in shaping research directions (highest eigenvector centrality = 1.00). Conclusions: The study revealed dynamic development of the physical culture discipline, with growing emphasis on interdisciplinarity and health aspects. Adaptation of research topics to current social and technological challenges was observed. Results indicate the need for further expansion of the thematic scope of theses, particularly in areas of new technologies and long-term effects of physical activity. The analysis provides valuable information for strategic planning of study program development and research directions in the field of physical culture, tourism, and recreation.

The impact of COVID-19 on private healthcare service utilisation: time series analysis in the capital region of Finland during 2020–2022

BackgroundMost current studies on COVID-19 have focused on the first weeks or months of the pandemic or have addressed specific diseases. The long-term effects of the COVID-19 pandemic on healthcare services are insufficiently understood. We analysed the long-term effects of the COVID-19 pandemic on outpatient healthcare utilisation in the full spectrum of diseases in Uusimaa, the capital region of Finland.MethodsOur data included 632 466 individual patients between the ages of 18 and 65 and 6 521 394 visits to healthcare personnel from electronic health records. We fitted an autoregressive integrated moving average (ARIMA) model to pre-2020 data to predict the expected quantity of diagnoses for the period between 1 January 2020 and 16 June 2022. Expected and observed quantities of diagnoses were compared.ResultsThe overall quantity of diagnoses declined by one-fourth immediately following the onset of the pandemic and remained around 15% below predicted values for most of the pandemic. After the pandemic-related restrictions were lifted, the total diagnostic activity started to recover to pre-pandemic projection levels. However, this recovery has been mainly driven by upper respiratory system-related activity. The number of diagnoses for many diagnosis groups has remained below the predicted levels, even after the concurrent removal of mobility restrictions and increased coverage of vaccinations in this population.ConclusionsThe pandemic resulted in an overall reduction in outpatient healthcare utilisation which persisted for 30 months. While the overall diagnostic activity has eventually recovered to predicted levels, many classes of diagnoses display reduced levels in the study population over the follow-up period. Some diseases that may have long-lasting effects when left untreated seem to remain underdiagnosed, potentially increasing pressure on the healthcare system in the future.

Time Series Analysis of Dry Chilli Price Trends in Byadagi Market, Haveri District, Karnataka, India

Spices are conventional aromatic vegetables mainly utilized for flavoring of food. Among these, chilli (capsicum annuum), is one of most important spice used around the world. The present study examined dry chillies price behaviour in the Byadagi market of Haveri District, Karnataka. The study was based on secondary data from 2000-2022 collected in Agmarketnet.in. Time series analysis involves a complex mix of four main components: trend (Tt), seasonal (St), cyclical (Ct), and irregular (It) variations. The result revealed that there was a significant increasing trend in prices of dry chillies from 2000-2022. The Multiplicative model of TCSI was used in this study, the polynomial regression was also used for the trend analysis and found the Third-degree model best fit. The highest seasonal index was found in January (112.44) while the lowest seasonal index in the month of July (93.28) the analysis of seasonal indices in the price of dry chilli would help to understand the status of dry chilli marketing in the Byadagi market. The dry chilli prices exhibit no oscillatory movement and no definite periodicity during the study period.

In Case You Haven't Heard…

Mondays and New Year's Day have the highest suicide risk according to new research, “Association of holidays and the day of the week with suicide risk: multi‐country, two‐stage, time series study,” published in The BMJ Oct. 23, according to a news release. An international research team used an advanced statistical technique called standardized two‐stage time series analysis to investigate suicide risk patterns. They analyzed data from 740 locations across 26 countries, gathered from the Multi‐country Multi‐city Collaborative Research Network database. This dataset comprised 1.7 million suicide cases, including daily suicide counts along with daily mean temperature data spanning from January 1971 to December 2019. The results revealed that suicide risk was highest on Mondays across all countries, though the effect of weekends varied. New Year's Day was associated with an increase in suicide risk in all countries, while Christmas showed a mixed pattern. Although there was an overall decreasing trend in suicide risk on other national holidays, the risk tended to increase after these holidays in most countries. Notably, the findings also showed that men are more vulnerable to variations related to the day of the week and New Year's Day compared to women. Researchers hope that this large‐scale study can lead to timely interventions, potentially saving countless lives.

Integrated time-series analysis of remote sensing imagery and crowdsourced data to monitor water hyacinth (Eichhornia crassipes) in the Orontes (asi) River

ABSTRACT Invasive plants are a serious global issue due to the damage they cause to aquatic ecosystems. In recent years, the Orontes River has been struggling with water hyacinth, an invasive species that completely covers the river’s surface. Within the scope of this study, the history of water hyacinth in the Orontes River was examined using remote sensing and crowdsourced data. Optical data with different spatial resolutions (Landsat-8 OLI and Sentinel-2 MSI) and Sentinel-1 SAR satellite data were used as remote sensing data. Google News, used as crowdsourced data, was compared with time series obtained from satellite data, emphasizing the complementary role of all three data types (optical, SAR and crowdsourced). Using PlanetScope Dove satellite data, the growth, development and extinction cycle of water hyacinth within a year were investigated according to water hyacinth density. In line with the results obtained, it is revealed that the water hyacinth began to appear after June, grew in the summer months, reached its peak in October–November, and disappeared after November with the increasing precipitation. This study is the first to examine water hyacinths in the Orontes River using remote sensing and crowdsourced data, and the results will guide decision-makers in the fight against water hyacinths.

Dynamic pricing and demand forecasting: Integrating time-series analysis, regression models, machine learning, and competitive analysis

Abstract. Dynamic pricing is a critical strategy for businesses seeking to optimize revenue and stay competitive in fluctuating markets. This paper explores the integration of various demand forecasting techniques, including time-series analysis, regression models, and machine learning algorithms, with competitive analysis methodologies to enhance dynamic pricing strategies. Time-series analysis focuses on decomposing data into trend, seasonality, and random fluctuations, using ARIMA models for accurate demand prediction. Regression models delve into the complexities of variable interactions, extending beyond linear relationships to include advanced techniques like Ridge, Lasso, and Elastic Net regression. Machine learning algorithms, such as decision trees, random forests, gradient boosting, and neural networks, revolutionize demand forecasting by uncovering complex patterns in large datasets. Competitive analysis incorporates market scanning, price elasticity estimation, and competitor behavior modeling to inform dynamic pricing decisions. Optimization algorithms, including linear programming, genetic algorithms, and simulated annealing, are employed to refine pricing strategies, while revenue management techniques like yield management and overbooking ensure maximum revenue from perishable goods and services. This comprehensive approach enables businesses to dynamically adjust prices in real-time, maintaining competitiveness and maximizing profitability in an ever-evolving market landscape.

Assessing the Carbon Intensity of e-fuels Production in European Countries: A Temporal Analysis

The transport sector heavily relies on the use of fossil fuels, which are causing major environmental concerns. Solutions relying on the direct or indirect use of electricity through e-fuel production are emerging to power the transport sector. To ensure environmental benefits are achieved over this transition, an accurate estimation of the impact of the use of electricity is needed. This requires a high temporal resolution to capture the high variability of electricity. This paper presents a previously unseen temporal analysis of the carbon intensity of e-fuels using grid electricity in countries that are members of the European Network of Transmission System Operators (ENTSO-E). It also provides an estimation of the potential load factor for producing low-carbon e-fuels according to the European Union legislative framework. This was achieved by building on top of the existing EcoDynElec tool to develop EcoDynElec_xr, a python tool enabling—with an hourly time resolution—the calculation, visualisation, and analysis of the historical time-series of electricity mixing from the ENTSO-E. The results highlight that, in 2023, very few European countries were reaching low carbon intensity for electricity that enables the use of grid electricity for the production of green electrolytic hydrogen. The methodological assumptions consider the consumption of the electricity mix instead of the production mix, and the considered time step is of paramount importance and drastically impacts the potential load factor of green hydrogen production. The developed tools are released under an open-source license to ensure transparency, result reproducibility, and reuse regarding newer data for other territories or for other purposes.

FORECASTING FIRE RISK VALUES USING MACHINE LEARNING AND TIME SERIES ANALYSIS

The probability of occurrence of emergencies and their consequences can be minimized by predicting events that contribute to the emergence of such a situation. An optimal and useful tool for such prediction and forecasting can be modeling, which can be carried out on the basis of information technology using computer programs.The methods of modeling and forecasting emergencies related to fires are the same as for other emergencies, but there are some specifics. The forecast of emergencies related to fires is aimed at three main stages: 1. forecasting the probability of occurrence with the determination of the place of the highest probability of fire occurrence, time and factors contributing to it; 2. forecasting the development of the situation and the spread of the fire (direction, speed) when it occurs, the process itself; 3. forecasting the consequences of a fire at a specific facility.The forecast is made based on the analysis of all possible unfavorable factors, their fields of action, features of the facility, its fire hazard and other data.The use of machine learning methods and time series analysis for forecasting fire risk values is substantiated. The possibility of using statistical methods of mathematical modeling of fire occurrence as the basis of the machine learning model is noted. This approach enables the machine learning model to process and analyze a large amount of statistical information for specific time periods and issue forecasts.

Ensemble deep learning techniques for time series analysis: a comprehensive review, applications, open issues, challenges, and future directions

Ensemble deep learning techniques for time series analysis: a comprehensive review, applications, open issues, challenges, and future directions