A probabilistic framework for identifying anomalies in urban air quality data.

Just as the value of crude oil is unlocked through refining, the true potential of air quality data is realized through systematic processing, analysis, and application. This refined data is critical for making informed decisions that may protect health and the environment. Perhaps ground-based air quality monitoring data often face quality control issues, notably outliers. The outliers in air quality data are reported as error and event-based. The error-based outliers are due to instrument failure, self-calibration, sensor drift over time, and the event based focused on the sudden change in meteorological conditions. The event-based outliers are meaningful while error-based outliers are noise that needs to be eliminated and replaced post-detection. In this study, we address error-based outlier detection in air quality data, particularly targeting particulate pollutants (PM2.5 and PM10) across various monitoring sites in Delhi. Our research specifically examines data from sites with less than 5% missing values and identifies four distinct types of error-based outliers: extreme values due to measurement errors, consecutive constant readings and low variance due to instrument malfunction, periodic outliers from self-calibration exceptions, and anomalies in the PM2.5/PM10 ratio indicative of issues with the instruments' dryer unit. We developed a robust methodology for outlier detection by fitting a non-linear filter to the data, calculating residuals between observed and predicted values, and then assessing these residuals using a standardized Z-score to determine their probability. Outliers are flagged based on a probability threshold established through sensitivity testing. This approach helps distinguish normal data points from suspicious ones, ensuring the refined quality of data necessary for accurate air quality modeling. This method is essential for improving the reliability of statistical and machine learning models that depend on high-quality environmental data.

Integration of Mitoflash and Time-Series Transcriptomics Facilitates Energy Dynamics Tracking and Substrate Supply Analysis of Floral Thermogenesis in Lotus.

The high biosynthetic and energetic demands of floral thermogenesis render thermogenic plants the ideal systems to characterize energy metabolism in plants, but real-time tracking of energy metabolism in plant cells remains challenging. In this study, a new method was developed for tracking the mitochondrial energy metabolism at the single mitochondria level by real-time imaging of mitochondrial superoxide production (i.e., mitoflash). Using this method, we observed the increased mitoflash frequencies in the receptacles of Nelumbo nucifera Gaertn. at the thermogenic stages. This increase, combined with the higher expression of antioxidant response-related genes identified through time-series transcriptomics at the same stages, shows us a new regulatory mechanism for plant redox balance. Furthermore, we found that the upregulation of respiratory metabolism-related genes during the thermogenic stages not only correlates with changes in mitoflash frequency but also underscores the critical roles of these pathways in ensuring adequate substrate supply for thermogenesis. Metabolite analysis revealed that sugars are likely one of the substrates for thermogenesis and may be transported over long distances by sugar transporters. Taken together, our findings demonstrate that mitoflash is a reliable tool for tracking energy metabolism in thermogenic plants and contributes to our understanding of the regulatory mechanisms underlying floral thermogenesis.

Heatwave and upper urinary tract stones morbidity: effect modification by heatwave definitions, disease subtypes, and vulnerable populations.

As heatwave occurs with increased frequency and intensity, the disease burden for urolithiasis, a heat-specific disease, will increase. However, heatwave effect on urolithiasis subtypes morbidity and optimal heatwave definition for urolithiasis remain unclear. Distributed lagged linear models were used to assess the associations between 32 defined heatwave and upper urinary tract stones morbidity. Relative risk (RR) and attributable fraction (AF) of upper urinary tract stone morbidity associated with heatwave of different intensities (low, middle, and high) were pooled by meta-analysis. Optimal heatwave definition was selected based on the combined score of AF, RR, and quasi-Akaike Information Criterion (QAIC) value. Stratified analyses were conducted to investigate the modification effects of gender, age, and disease subtypes. Association between heatwave and upper urinary tract stones morbidity was mainly for ureteral calculus, and AF was highest for low-intensity heatwave. This study's optimal heatwave was defined as average temperature > 93rd percentile for ≥ 2 consecutive days, with AF of 7.40% (95% CI: 2.02%, 11.27%). Heatwave was associated with ureteral calculus morbidity in males and middle-aged adults. While heatwave effect was statistically insignificant in females and other age groups. Managers should develop appropriate definitions to address heatwave based on regional characteristics and focus on heatwave effects on urolithiasis.

Effect of time series on the degradation of lignin by Trametes gibbosa: Products and pathways.

Lignin is the third most abundant organic resource in nature. The utilization of white-rot fungi for wood degradation effectively circumvents environmental pollution associated with chemical treatments, facilitating the benign decomposition of lignin. Trametes gibbosa is a typical white-rot fungus with rapid growth and strong wood decomposition ability. The lignin content decreased from 23.62 mg/mL to 17.05 mg/mL, which decreased by 27 % in 30 days. The activity of manganese peroxidase increased steadily by 9.44 times. The activities of laccase and lignin peroxidase had the same trend of change and reached peaks of 49.88 U/L and 10.43 U/L on the 25th day, respectively. The change in H2O2 content in vivo was opposite to its trend. For FTIR and GC-MS analysis, the fungi attacked the side chain structure of lignin phenyl propane polymer and benzene ring to crack into low molecular weight aromatic compounds. The side chains of low molecular weight aromatic compounds are oxidized, and long-chain carboxylic acids are formed. Additionally, the absorption peak in the vibration region of the benzene ring skeleton became complex, and the structure of the benzene rings changed. In the beginning, fungal growth was inhibited. Fungal autophagy was aggravated. The metal cation binding proteins of fungi were active, and the genes related to detoxification metabolism were upregulated. The newly produced compounds are related to xenobiotic metabolism. The degradation peak focused on the redox process, and the biological function was enriched in the regulation of macromolecular metabolism, lignin metabolism, and oxidoreductase activity acting on diphenols and related substances as donors. Notably, genes encoding key degradation enzymes, including lcc3, lcc4, phenol-2-monooxygenase, 3-hydroxybenzoate-6-hydroxylase, oxalate decarboxylase, and acetyl-CoA oxidase were significantly upregulated. On the 30th day, the N-glycan biosynthesis pathway was significantly enriched in glycan biosynthesis and metabolism. Weighted correlation network analysis was performed. A total of 1452 genes were clustered in the coral1 module, which were most related to lignin degradation. The genes were significantly enriched in oxidoreductase activity, peptidase activity, cell response to stimulation, signal transduction, lignin metabolism, and phenylpropane metabolism, while the rest were concentrated in glucose metabolism. In this study, the lignin degradation process and products were revealed by T. gibbosa. The molecular mechanism of lignin degradation in different stages was explored. The selection of an efficient utilization time of lignin will help to increase the degradation rate of lignin. This study provides a theoretical basis for the biofuel and biochemical production of lignin. SYNOPSIS: Trametes gibbosa degrades lignin in a pollution-free way, improving the utilization of carbon resources in an environmentally friendly spontaneous cycle. The products are the new way towards sustainable development and low-carbon technology.

Application of time series analysis in predicting postpartum depression: integrating data from the hospitalization period and early postpartum weeks.

PURPOSE: This study aimed to explore the dynamic changes in postpartum depressive symptoms from the hospitalization period to 4-8 weeks postpartum using time series analysis techniques. By integrating depressive scores from the hospital stay and the early postpartum weeks, we sought to develop a predictive model to enhance early identification and intervention strategies for Postpartum Depression (PPD). METHODS: A longitudinal design was employed, analyzing Edinburgh Postnatal Depression Scale (EPDS) scores from 1,287 postpartum women during hospitalization and at 4, 6, and 8 weeks postpartum. Descriptive statistics summarized demographic characteristics. Time Series Analysis using the Auto-Regressive Integrated Moving Average (ARIMA) model explored temporal trends and seasonal variations in EPDS scores. Correlation analysis examined the relationships between EPDS scores and demographic characteristics. Model validation was conducted using a separate dataset. RESULTS: EPDS scores significantly increased from the hospitalization period to 4-8 weeks postpartum (p < .001). The ARIMA model revealed seasonal and trend variations, with higher depressive scores in the winter months. The model's fit indices (AIC = 765.47; BIC = 774.58) indicated a good fit. The Moving Average (MA) coefficient was - 0.69 (p < .001), suggesting significant negative impacts from previous periods' errors. CONCLUSIONS: Monitoring postpartum depressive symptoms dynamically was crucial, particularly during the 4-8 weeks postpartum. The seasonal trend of higher depressive scores in winter underscored the need for tailored interventions. Further research using longitudinal and multi-center designs was warranted to validate and extend these findings. Our predictive model aimed to enhance early identification and intervention strategies, contributing to better maternal and infant health outcomes.

Short- and medium-term cumulative effects of traffic-related air pollution on resting heart rate in the elderly: A wearable device study.

BACKGROUND: Epidemiological evidence regarding the association between air pollution and resting heart rate (RHR), a predictor of cardiovascular disease and mortality, is limited and inconsistent. OBJECTIVES: We used wearable devices and time-series analysis to assess the exposure-response relationship over an extended lag period. METHODS: Ninety-seven elderly individuals (>65 years) from the Taipei Basin participated from May to November 2020 and wore Garmin® smartwatches continuously until the end of 2021 for heart rate monitoring. RHR was defined as the daily average of the lowest 30-min heart rate. Air pollution exposure data, covering lag periods from 0 to 60 days, were obtained from nearby monitoring stations. We used distributed lag non-linear models and linear mixed-effect models to assess cumulative effects of air pollution. Principal component analysis was utilized to explore underlying patterns in air pollution exposure, and subgroup analyses with interaction terms were conducted to explore the modification effects of individual factors. RESULTS: After adjusting for co-pollutants in the models, an interquartile range increase of 0.18 ppm in carbon monoxide (CO) was consistently associated with increased RHR across lag periods of 0-1 day (0.31, 95 % confidence interval [CI]: 0.24-0.38), 0-7 days (0.68, 95 % CI: 0.57-0.79), and 0-50 days (1.02, 95 % CI: 0.82-1.21). Principal component analysis identified two factors, one primarily influenced by CO and nitrogen dioxide (NO2), indicative of traffic sources. Increases in the varimax-rotated traffic-related score were correlated with higher RHR over 0-1 day (0.36, 95 % CI: 0.25-0.47), 0-7 days (0.62, 95 % CI: 0.46-0.77), and 0-50 days (1.27, 95 % CI: 0.87-1.67) lag periods. Over a 0-7 day lag, RHR responses to traffic pollution were intensified by higher temperatures (ß = 0.80 vs. 0.29; interaction p-value [P_int] = 0.011). Males (ß = 0.66 vs. 0.60; P_int < 0.0001), hypertensive individuals (ß = 0.85 vs. 0.45; P_int = 0.028), diabetics (ß = 0.96 vs. 0.52; P_int = 0.042), and those with lower physical activity (ß = 0.70 vs. 0.54; P_int < 0.0001) also exhibited stronger responses. Over a 0-50 day lag, males (ß = 0.99 vs. 0.96; P_int < 0.0001), diabetics (ß = 1.66 vs. 0.69; P_int < 0.0001), individuals with lower physical activity (ß = 1.49 vs. 0.47; P_int = 0.0006), and those with fewer steps on lag day 1 (ß = 1.17 vs. 0.71; P_int = 0.029) showed amplified responses. CONCLUSIONS: Prolonged exposure to traffic-related air pollution results in cumulative cardiovascular risks, persisting for up to 50 days. These effects are more pronounced on warmer days and in individuals with chronic conditions or inactive lifestyles.

The relationship of short-term exposure to meteorological factors on diabetes mellitus mortality risk in Hefei, China: a time series analysis.

OBJECTIVES: The study aims to explore whether short-term exposure to meteorological factors has a potential association with the risk of diabetes mellitus (DM) mortality. METHODS: During the period 2015-2018, we collected daily data on meteorological factors and deaths of diabetic patients in Hefei. A total of 1101 diabetic deaths were recorded. We used structural equation modeling to initially explore the relationships among air pollutants, meteorological variables, and mortality, and generalized additive modeling (GAM) and distributional lag nonlinear modeling (DLNM) to explore the relationship between meteorological factors and the mortality risk of DM patients. We also stratified by age and gender. The mortality risk in diabetic patients was expressed by relative risks (RR) and 95% confidence intervals (CI) for both single and cumulative days. RESULTS: Single-day lagged results showed a high relative humidity (RH) (75th percentile, 83.71%), a fairly high average temperature (T mean) (95th percentile, 30.32 °C), and an extremely low diurnal temperature range (DTR) (5th percentile, 3.13 °C) were positively related to the mortality risk of DM. Stratified results showed that high and very high levels of T mean were significantly positively linked to the mortality risk of DM among females and the elderly, while very high levels of DTR were linked to the mortality risk in men and younger populations. CONCLUSION: In conclusion, this study found that short-duration exposure to quite high T mean, high RH, and very low DTR were significantly positively related to the mortality risk of DM patients. For women and older individuals, exposure to high and very high T mean environments should be minimized. Men and young adults should be aware of daily temperature changes.

An analytical approach for identifying trend-seasonal components and detecting unexpected behaviour in psychological time-series.

The recent advances in technological capabilities have led to a massive production of time-series data and remarkable progress in longitudinal designs and analyses within psychological research. However, implementing time-series analysis can be challenging due to the various characteristics and complexities involved, as well as the need for statistical expertise. This paper introduces a statistical pipeline on time-series analysis for studying the changes in a single process over time at either a population or individual level, both retrospectively and prospectively. This is achieved through systemization and extension of existing modelling and inference techniques. This analytical approach enables practitioners not only to track but also to model and evaluate emerging trends and apparent seasonality. It also allows for the detection of unexpected events, quantifying their deviations from baseline and forecasting future values. Given that other discernible population- and individual-level changes in psychological and behavioural processes have not yet emerged, continued surveillance is warranted. A near real-time monitoring tool of time-series data could guide community psychological responses across multiple ecological levels, making it a valuable resource for field practitioners and psychologists. An empirical study is conducted to illustrate the implementation of the introduced analytical pipeline in practice and to demonstrate its capabilities.

Changes in Mental Health Among Adolescents in South Korea Before and After COVID-19: An Interrupted Time Series Analysis From 2015 to 2022.

PURPOSE: This study aimed to examine changes in mental health among adolescents by comparing data from the period following the onset of the COVID-19 pandemic with the period before the pandemic. METHODS: We estimated the annual prevalence of stress perception, depressive symptoms, and suicidal ideation among middle and high school students using data from the Korean Youth Health Behavior Survey spanning from 2015 to 2022. We then compared mental health status across 2 periods-pre-COVID-19 (2015-2019) and during COVID-19 (2020-2022)-employing an interrupted time series analysis. We adjusted for covariates, such as household economic status, residence type, self-rated health, and history of hospitalization, due to violence. RESULTS: We analyzed data from 472,385 adolescents (242,819 boys and 230,016 girls). Stress perception, depressive symptoms, and suicidal ideation showed an increasing trend during the pre-COVID-19 period, followed by a decrease in the first year of the pandemic and an increasing trend in the second and third years. Boys experienced a faster increase in stress and depressive symptoms during the second and third years of the pandemic compared with the pre-COVID-19 period, whereas girls showed trends similar to those observed before the pandemic. Middle school students experienced a more rapid increase in these indicators than high school students during the second and third years. DISCUSSION: Adolescents' mental health initially improved in the first year of COVID-19 but worsened during the second and third years of the pandemic. This suggests a need for intervention policies and programs to support adolescent mental health.

Oxytocin differentially modulates reward system responses to social and non-social incentives.

RATIONALE: Oxytocin has been shown to modulate behavior related to processing of monetary incentives and to regulate social and reproductive behavior, yet little is known about how oxytocin differentially influences neural responses to social and non-social incentives. OBJECTIVES: We aimed to evaluate the effects of oxytocin administration on behavioral and neural responses to social and monetary incentives. METHODS: Twenty-eight healthy adults (age 18-45 years) performed both monetary and social incentive tasks during blood oxygenation level dependent (BOLD) imaging. Intranasal oxytocin or placebo was administered before each scan using a double blind, randomized, cross-over design. Task performance and self-reported motivation and mood states were collected. Time-series analysis was conducted to assess the influence of oxytocin on the hemodynamic response in the ventral tegmental area and substantia nigra (VTA/SN) and nucleus accumbens (NAc). RESULTS: Oxytocin demonstrated a multifaceted effect on VTA/SN and NAc when processing reward incentives, with it increasing BOLD response in VTA/SN and decreasing BOLD response in NAc during social incentive anticipation. A reversal of this was shown with decreased BOLD responses in the VTA/SN and increased BOLD response in the NAc during monetary incentive anticipation. CONCLUSIONS: Our findings suggest a more nuanced purpose of oxytocin when evaluating reward incentive decision making. It is possible that while oxytocin does increase salience to rewards, that it is more important for cognitive control when determining short-term versus long-term benefits in rewards. Future studies should more closely examine the relationship between oxytocin and delay discounting.

Characterization of the reproductive strategy of invasive Round Goby (Neogobius melanostomus) in the Upper Danube River.

Originating from the Black and Caspian seas, the Round Goby (Neogobius melanostomus) has become one of the most successful invaders of freshwater ecosystems. In this study, we provide a characterization of the reproductive strategy of an established population of Round Gobies in the Upper Danube river including sex ratio, fluctuations of gonadosomatic index (GSI), analysis of timing of spawning as well as of clutch and egg size. We compare these results to other studies from the native and invaded range. In the Danube, the Round Goby population was found to be female dominated, however fluctuations in magnitude of female bias were observed between months. Monitoring of the population across 1.5 years revealed that GSI was highest from April to June, while lowest values were observed in August and September. Using time-series analysis, a delayed effect of temperature on GSI was found for females and males, while a quicker response of GSI levels to photoperiod and discharge was observed for females. GSI increased with body size for females and eggs were found to be significantly larger in May, however clutch sizes did not differ between months. Results of a literature review revealed great differences in timing and length of spawning season as well as sex ratio between populations throughout the distribution range, which can probably be explained by climatic and photoperiodic conditions together with the time since invasion and the high plasticity of Round Gobies.

Transferability of ecological forecasting models to novel biotic conditions in a long-term experimental study.

Ecological forecasting models play an increasingly important role for managing natural resources and assessing our fundamental knowledge of processes driving ecological dynamics. As global environmental change pushes ecosystems beyond their historical conditions, the utility of these models may depend on their transferability to novel conditions. Because species interactions can alter resource use, timing of reproduction, and other aspects of a species' realized niche, changes in biotic conditions, which can arise from community reorganization events in response to environmental change, have the potential to impact model transferability. Using a long-term experiment on desert rodents, we assessed model transferability under novel biotic conditions to better understand the limitations of ecological forecasting. We show that ecological forecasts can be less accurate when the models generating them are transferred to novel biotic conditions and that the extent of model transferability can depend on the species being forecast. We also demonstrate the importance of incorporating uncertainty into forecast evaluation with transferred models generating less accurate and more uncertain forecasts. These results suggest that how a species perceives its competitive landscape can influence model transferability and that when uncertainties are properly accounted for, transferred models may still be appropriate for decision making. Assessing the extent of the transferability of forecasting models is a crucial step to increase our understanding of the limitations of ecological forecasts.

Predicting herd immunity achievement: a time-series analysis of vaccination and fatality rates using 1,075 days of COVID-19 data.

Introduction: The COVID-19 pandemic, driven by SARS-CoV-2, has made vaccination a critical strategy for global control. However, vaccine hesitancy, particularly among certain age groups, remains a significant barrier to achieving herd immunity. Methods: This study uses Poisson regression and ARIMA time-series modeling to identify factors contributing to vaccine hesitancy, understand age-specific vaccination preferences, and assess the impact of bivalent vaccines on reducing hesitancy and fatality rates. It also predicts the time required to achieve herd immunity by analyzing factors such as vaccine dosing intervals, age-specific preferences, and changes in fatality rates. Results: The study finds that individuals recovering from COVID-19 often delay vaccination due to perceived immunity. There is a preference for combining BNT162b2 and CoronaVac vaccines. The BNT162b2 bivalent vaccine has significantly reduced vaccine hesitancy and is linked with lower fatality rates, particularly in those aged 80 and above. However, it tends to induce more severe side effects compared to Sinovac. Vaccine hesitancy is most prevalent among the youngest (0-11) and oldest (80+) age groups, posing a challenge to reaching 90% vaccination coverage. Conclusion: Vaccine hesitancy is a major obstacle to herd immunity. Effective strategies include creating urgency, offering incentives, and prioritizing vulnerable age groups. Despite these challenges, the government should have continued to encourage vaccinations while gradually lifting COVID-19 control measures, balancing public health safety with the return to normal life, as was observed in the transition period during the latter stages of the pandemic.

A global model-agnostic rule-based XAI method based on Parameterized Event Primitives for time series classifiers.

Time series classification is a challenging research area where machine learning and deep learning techniques have shown remarkable performance. However, often, these are seen as black boxes due to their minimal interpretability. On the one hand, there is a plethora of eXplainable AI (XAI) methods designed to elucidate the functioning of models trained on image and tabular data. On the other hand, adapting these methods to explain deep learning-based time series classifiers may not be straightforward due to the temporal nature of time series data. This research proposes a novel global post-hoc explainable method for unearthing the key time steps behind the inferences made by deep learning-based time series classifiers. This novel approach generates a decision tree graph, a specific set of rules, that can be seen as explanations, potentially enhancing interpretability. The methodology involves two major phases: (1) training and evaluating deep-learning-based time series classification models, and (2) extracting parameterized primitive events, such as increasing, decreasing, local max and local min, from each instance of the evaluation set and clustering such events to extract prototypical ones. These prototypical primitive events are then used as input to a decision-tree classifier trained to fit the model predictions of the test set rather than the ground truth data. Experiments were conducted on diverse real-world datasets sourced from the UCR archive, employing metrics such as accuracy, fidelity, robustness, number of nodes, and depth of the extracted rules. The findings indicate that this global post-hoc method can improve the global interpretability of complex time series classification models.

Assessing the impact of the COVID-19 pandemic on the mental health-related hospitalization rate of youth in Canada: an interrupted time series analysis. / Évaluation de l'effet de la pandémie de COVID-19 sur le taux d'hospitalisation pour des raisons de santé mentale chez les jeunes au Canada : analyse de séries chronologiques interrompues.

INTRODUCTION: This study evaluated the effect of the COVID-19 pandemic on temporal trends in mental health and addiction-related inpatient hospitalization rates among youth (aged 10-17 years) in Canadian provinces and territories (excluding Quebec) from 1 April 2018 to 5 March 2022. METHODS: We conducted an interrupted time series analysis across three periods: T0 (pre-pandemic: 1 April 2018 to 15 March 2020); T1 (early pandemic: 15 March 2020 to 5 July 2020); and T2 (later pandemic: 6 July 2020 to 5 March 2022). RESULTS: Pre-pandemic mental health and addiction-related hospitalization rates had significant regional variability, with weekly rates from 6.27 to 85.59 events per 100 000 persons in Manitoba and the territories combined, respectively. During T1, the national (excluding Quebec) weekly hospitalization rate decreased from a pre-pandemic level of 12.82 (95% CI: 12.14 to 13.50) to 5.11 (95% CI: 3.80 to 6.41) events per 100 000 persons. There was no statistically significant change in the mental health and addiction- related hospitalization rate across provinces and territories in T2 compared to T0. However, there was a significant increase in the rate of self-harm-related hospitalizations among females Canada-wide and in most provinces during this period. CONCLUSION: Although several Canadian studies have reported increases in mental health and addiction-related outpatient and emergency department visits among youth during the COVID-19 pandemic, this did not correspond to an increase in the inpatient hospital burden, with the notable exception of self-harm among young females.

LSTM networks provide efficient cyanobacterial blooms forecasting even with incomplete spatio-temporal data.

Cyanobacteria are the most frequent dominant species of algal blooms in inland waters, threatening ecosystem function and water quality, especially when toxin-producing strains predominate. Enhanced by anthropogenic activities and global warming, cyanobacterial blooms are expected to increase in frequency and global distribution. Early Warning Systems (EWS) for cyanobacterial blooms development allow timely implementation of management measures, reducing the risks associated to these blooms. In this paper, we propose an effective EWS for cyanobacterial bloom forecasting, which uses 6 years of incomplete high-frequency spatio-temporal data from multiparametric probes, including phycocyanin (PC) fluorescence as a proxy for cyanobacteria. A probe agnostic and replicable method is proposed to pre-process the data and to generate time series specific for cyanobacterial bloom forecasting. Using these pre-processed data, six different non-site/species-specific predictive models were compared including the autoregressive and multivariate versions of Linear Regression, Random Forest, and Long-Term Short-Term (LSTM) neural networks. Results were analyzed for seven forecasting time horizons ranging from 4 to 28 days evaluated with a hybrid system that combined regression metrics (MSE, R2, MAPE) for PC values, classification metrics (Accuracy, F1, Kappa) for a proposed alarm level of 10 µg PC/L, and a forecasting-specific metric to measure prediction improvement over the displaced signal (skill). The multivariate version of LSTM showed the best and most consistent results across all forecasting horizons and metrics, achieving accuracies of up to 90 % in predicting the proposed PC alarm level. Additionally, positive skill values indicated its outstanding effectiveness to forecast cyanobacterial blooms from 16 to 28 days in advance.

Exploring the influence of environmental indicators and forecasting influenza incidence using ARIMAX models.

Background: Influenza is a respiratory infection that poses a significant health burden worldwide. Environmental indicators, such as air pollutants and meteorological factors, play a role in the onset and propagation of influenza. Accurate predictions of influenza incidence and understanding the factors influencing it are crucial for public health interventions. Our study aims to investigate the impact of various environmental indicators on influenza incidence and apply the ARIMAX model to integrate these exogenous variables to enhance the accuracy of influenza incidence predictions. Method: Descriptive statistics and time series analysis were employed to illustrate changes in influenza incidence, air pollutants, and meteorological indicators. Cross correlation function (CCF) was used to evaluate the correlation between environmental indicators and the influenza incidence. We used ARIMA and ARIMAX models to perform predictive analysis of influenza incidence. Results: From January 2014 to September 2023, a total of 21,573 cases of influenza were reported in Fuzhou, with a noticeable year-by-year increase in incidence. The peak of influenza typically occurred around January each year. The results of CCF analysis showed that all 10 environmental indicators had a significant impact on the incidence of influenza. The ARIMAX(0, 0, 1) (1, 0, 0)12 with PM10(lag5) model exhibited the best prediction performance, as indicated by the lowest AIC, AICc, and BIC values, which were 529.740, 530.360, and 542.910, respectively. The model achieved a fitting RMSE of 2.999 and a predicting RMSE of 12.033. Conclusion: This study provides insights into the impact of environmental indicators on influenza incidence in Fuzhou. The ARIMAX(0, 0, 1) (1, 0, 0)12 with PM10(lag5) model could provide a scientific basis for formulating influenza control policies and public health interventions. Timely prediction of influenza incidence is essential for effective epidemic control strategies and minimizing disease transmission risks.

MLFGCN: short-term residential load forecasting via graph attention temporal convolution network.

Introduction: Residential load forecasting is a challenging task due to the random fluctuations caused by complex correlations and individual differences. The existing short-term load forecasting models usually introduce external influencing factors such as climate and date. However, these additional information not only bring computational burden to the model, but also have uncertainty. To address these issues, we propose a novel multi-level feature fusion model based on graph attention temporal convolutional network (MLFGCN) for short-term residential load forecasting. Methods: The proposed MLFGCN model fully considers the potential long-term dependencies in a single load series and the correlations between multiple load series, and does not require any additional information to be added. Temporal convolutional network (TCN) with gating mechanism is introduced to learn potential long-term dependencies in the original load series. In addition, we design two graph attentive convolutional modules to capture potential multi-level dependencies in load data. Finally, the outputs of each module are fused through an information fusion layer to obtain the highly accurate forecasting results. Results: We conduct validation experiments on two real-world datasets. The results show that the proposed MLFGCN model achieves 0.25, 7.58% and 0.50 for MAE, MAPE and RMSE, respectively. These values are significantly better than those of baseline models. Discussion: The MLFGCN algorithm proposed in this paper can significantly improve the accuracy of short-term residential load forecasting. This is achieved through high-quality feature reconstruction, comprehensive information graph construction and spatiotemporal features capture.

Diurnal temperature range and cardiopulmonary health in Taiwan: Evaluating impacts, thresholds, and vulnerable groups.

The health impacts of the diurnal temperature range (DTR), which may be affected by climate change, have received little attention. The objectives of this study were (1) to evaluate the association of DTR and cardiopulmonary outcomes, (2) to select the proper thresholds for a DTR warning system, and (3) to identify vulnerable groups. The weather and health records in Taiwan from 2000 to 2019, with a maximum DTR of 12.8 °C, were analyzed using generalized additive models. The health outcomes included cardiovascular (CVD) and respiratory disease (RD) categories and several sub-categories, such as ischemic heart disease, stroke, pneumonia, asthma, and chronic obstructive pulmonary disease. The results showed that the associations of DTR and cardiopulmonary outcomes were as significant as, and sometimes even stronger than, those of the daily maximum temperature and daily minimum apparent temperature in the warm and cold seasons, respectively. The significant association began at DTR of 6 °C, lower than previously reported. The identified DTR warning thresholds were 8.5 and 11 °C for the warm and cold seasons, respectively. DTR is statistically significantly associated with a 5-36% and a 9-20% increase in cardiopulmonary emergency and hospitalized cases in the warm season with a 1 °C increase above 8.5 °C, respectively. In the cold season, DTR is significantly associated with 7-41%, 4-30%, and 36-100% increases in cardiopulmonary emergency, hospitalized, and mortality with a 1 °C increase above 11 °C, respectively. People with hypertension, hyperglycemia, and hyperlipidemia had even higher risks. Vulnerable age and sex groups were identified if they had a lower DTR-health threshold than the general population, which can be integrated into a warning system. In conclusion, DTR may be increased on a local or city scale under climate change; a DTR warning system and vulnerable group identification may be warranted in most countries for health risk reduction.

Dynamics of following and leading: association of movement synchrony and depression severity.

Objective: Depression negatively affects interpersonal functioning and influences nonverbal behavior. Interpersonal theories of depression suggest that depressed individuals engage in behaviors that initially provoke others' support and reassurance, but eventually lead to rejection that may also be expressed nonverbally. Methods: This study investigated movement synchrony as a nonverbal indicator of support and rejection and its association with depression severity in a sample of depressed and healthy individuals. Semi-standardized diagnostic interview segments with N = 114 dyads were video recorded. Body movement was analyzed using Motion Energy Analysis, synchrony intervals were identified by computing windowed cross-lagged correlation and a peak-picking-algorithm. Depression severity was assessed via both self-rating (BDI-II) and clinician rating (HAMD). Results: Both self-rated and clinician-rated depression severity were negatively correlated with patient-led, but not clinician-led movement synchrony measures. The more depressed patients were, the less they initiated movement synchrony with their clinicians. These correlations remained significant after controlling for gender, age, gross body movement, and psychopharmacological medication. Conclusion: Findings suggest that depression may negatively affect patients' active initiative in interaction situations. Automatized methods as used in this study can add valuable information in the diagnosis of depression and the assessment of associated social impairments.