A regionally tailored epidemiological forecast and monitoring program to guide a healthcare system in the COVID-19 pandemic.

BACKGROUND: During the COVID-19 pandemic, analytics and predictive models built on regional data provided timely, accurate monitoring of epidemiological behavior, informing critical planning and decision-making for health system leaders. At Atrium Health, a large, integrated healthcare system in the southeastern United States, a team of statisticians and physicians created a comprehensive forecast and monitoring program that leveraged an array of statistical methods. METHODS: The program utilized the following methodological approaches: (i) exploratory graphics, including time plots of epidemiological metrics with smoothers; (ii) infection prevalence forecasting using a Bayesian epidemiological model with time-varying infection rate; (iii) doubling and halving times computed using changepoints in local linear trend; (iv) death monitoring using combination forecasting with an ensemble of models; (v) effective reproduction number estimation with a Bayesian approach; (vi) COVID-19 patients hospital census monitored via time series models; and (vii) quantified forecast performance. RESULTS: A consolidated forecast and monitoring report was produced weekly and proved to be an effective, vital source of information and guidance as the healthcare system navigated the inherent uncertainty of the pandemic. Forecasts provided accurate and precise information that informed critical decisions on resource planning, bed capacity and staffing management, and infection prevention strategies. CONCLUSIONS: In this paper, we have presented the framework used in our epidemiological forecast and monitoring program at Atrium Health, as well as provided recommendations for implementation by other healthcare systems and institutions to facilitate use in future pandemics.

Forecasting species' responses to climate change using space-for-time substitution.

To anticipate species' responses to climate change, ecologists have largely relied on the space-for-time-substitution (SFTS) approach. However, the hypothesis and its underlying assumptions have been poorly tested. Here, we detail how the efficacy of using the SFTS approach to predict future locations will depend on species' traits, the ecological context, and whether the species is declining or introduced. We argue that the SFTS approach will be least predictive in the contexts where we most need it to be: forecasting the expansion of the range of introduced species and the recovery of threatened species. We highlight how evaluating the underlying assumptions, along with improved methods, will rapidly advance our understanding of the applicability of the SFTS approach, particularly in the context of modelling the distribution of species.

Using neural ordinary differential equations to predict complex ecological dynamics from population density data.

Simple models have been used to describe ecological processes for over a century. However, the complexity of ecological systems makes simple models subject to modelling bias due to simplifying assumptions or unaccounted factors, limiting their predictive power. Neural ordinary differential equations (NODEs) have surged as a machine-learning algorithm that preserves the dynamic nature of the data (Chen et al. 2018 Adv. Neural Inf. Process. Syst.). Although preserving the dynamics in the data is an advantage, the question of how NODEs perform as a forecasting tool of ecological communities is unanswered. Here, we explore this question using simulated time series of competing species in a time-varying environment. We find that NODEs provide more precise forecasts than autoregressive integrated moving average (ARIMA) models. We also find that untuned NODEs have a similar forecasting accuracy to untuned long-short term memory neural networks and both are outperformed in accuracy and precision by empirical dynamical modelling . However, we also find NODEs generally outperform all other methods when evaluating with the interval score, which evaluates precision and accuracy in terms of prediction intervals rather than pointwise accuracy. We also discuss ways to improve the forecasting performance of NODEs. The power of a forecasting tool such as NODEs is that it can provide insights into population dynamics and should thus broaden the approaches to studying time series of ecological communities.

Enhancing hydrological modeling with transformers: a case study for 24-h streamflow prediction.

In this paper, we address the critical task of 24-h streamflow forecasting using advanced deep-learning models, with a primary focus on the transformer architecture which has seen limited application in this specific task. We compare the performance of five different models, including persistence, long short-term memory (LSTM), Seq2Seq, GRU, and transformer, across four distinct regions. The evaluation is based on three performance metrics: Nash-Sutcliffe Efficiency (NSE), Pearson's r, and normalized root mean square error (NRMSE). Additionally, we investigate the impact of two data extension methods: zero-padding and persistence, on the model's predictive capabilities. Our findings highlight the transformer's superiority in capturing complex temporal dependencies and patterns in the streamflow data, outperforming all other models in terms of both accuracy and reliability. Specifically, the transformer model demonstrated a substantial improvement in NSE scores by up to 20% compared to other models. The study's insights emphasize the significance of leveraging advanced deep learning techniques, such as the transformer, in hydrological modeling and streamflow forecasting for effective water resource management and flood prediction.

Construction of Prediction Model of Foodborne Disease Outbreaks and Its Trend Prediction - Guizhou Province, China, 2023-2025.

Objective: Foodborne diseases pose a significant public health concern globally. This study aims to analyze the correlation between disease prevalence and climatic conditions, forecast the pattern of foodborne disease outbreaks, and offer insights for effective prevention and control strategies and optimizing health resource allocation policies in Guizhou Province. Methods: This study utilized the χ2 test and four comprehensive prediction models to analyze foodborne disease outbreaks recorded in the Guizhou Foodborne Disease Outbreak system between 2012 and 2022. The best-performing model was chosen to forecast the trend of foodborne disease outbreaks in Guizhou Province, 2023-2025. Results: Significant variations were observed in the incidence of foodborne disease outbreaks in Guizhou Province concerning various meteorological factors (all P≤0.05). Among all models, the SARIMA-ARIMAX combined model demonstrated the most accurate predictive performance (RMSE: Prophet model=67.645, SARIMA model=3.953, ARIMAX model=26.544, SARIMA-ARIMAX model=26.196; MAPE: Prophet model=42.357%, SARIMA model=37.740%, ARIMAX model=15.289%, SARIMA-ARIMAX model=13.961%). Conclusion: The analysis indicates that foodborne disease outbreaks in Guizhou Province demonstrate distinct seasonal patterns. It is recommended to concentrate prevention efforts during peak periods. The SARIMA-ARIMAX hybrid model enhances the precision of monthly forecasts for foodborne disease outbreaks, offering valuable insights for future prevention and control strategies.

Emotions for a novel health food: Is there an impact bias and can it be reduced?

Eating decisions may be influenced by an impact bias, or the tendency to overestimate the intensity and duration of emotions for future events or outcomes. In this study, we tested the impact bias, among college students, for eating a novel health food - kale chips. We also examined how an emotion adaptation recall exercise influenced emotions and eating behavior. After reading about the health benefits of kale, college students (N = 80) reported their anticipated emotions (e.g., how intensely they would experience each emotion) for eating kale chips. Following a control (n = 40) or emotion adaptation (n = 40) writing exercise, they were asked to eat these chips. They then reported their experienced emotions (e.g., how intensely they experienced each emotion) as well as their intentions to eat kale in the future. Findings indicated that participants showed a negative impact bias in which they anticipated more intense negative emotions than they experienced. However, they showed an opposite effect for positive emotions, anticipating lower positive emotion than they experienced. Relative to the control group, those who did an emotion adaptation exercise experienced lower levels of disgust after eating kale chips and reported higher intentions to eat them in the future. Subsequent analyses of consumption revealed that those in the emotion adaptation condition also ate more of the kale chips. Findings suggest that affective forecasting theory may be a useful framework through which to develop and test ideas about emotions and eating in the context of novel health foods.

A hybrid deep learning approach to predict hourly riverine nitrate concentrations using routine monitored data.

With high-frequency data of nitrate (NO3-N) concentrations in waters becoming increasingly important for understanding of watershed system behaviors and ecosystem managements, the accurate and economic acquisition of high-frequency NO3-N concentration data has become a key point. This study attempted to use coupled deep learning neural networks and routine monitored data to predict hourly NO3-N concentrations in a river. The hourly NO3-N concentration at the outlet of the Oyster River watershed in New Hampshire, USA, was predicted through neural networks with a hybrid model architecture coupling the Convolutional Neural Networks and the Long Short-Term Memory model (CNN-LSTM). The routine monitored data (the river depth, water temperature, air temperature, precipitation, specific conductivity, pH and dissolved oxygen concentrations) for model training were collected from a nested high-frequency monitoring network, while the high-frequency NO3-N concentration data obtained at the outlet were not included as inputs. The whole dataset was separated into training, validation, and testing processes according to the ratio of 5:3:2, respectively. The hybrid CNN-LSTM model with different input lengths (1d, 3d, 7d, 15d, 30d) displayed comparable even better performance than other studies with lower frequencies, showing mean values of the Nash-Sutcliffe Efficiency 0.60-0.83. Models with shorter input lengths demonstrated both the higher modeling accuracy and stability. The water level, water temperature and pH values at monitoring sites were main controlling factors for forecasting performances. This study provided a new insight of using deep learning networks with a coupled architecture and routine monitored data for high-frequency riverine NO3-N concentration forecasting and suggestions about strategies about variable and input length selection during preprocessing of input data.

Voter emotional responses and voting behaviour in the 2020 US presidential election.

Political polarisation in the United States offers opportunities to explore how beliefs about candidates - that they could save or destroy American society - impact people's thoughts, feelings, and behaviour. Participants forecast their future emotional responses to the contentious 2020 U.S. presidential election, and reported their actual responses after the election outcome. Stronger beliefs about candidates were associated with forecasts of greater emotion in response to the election, but the strength of this relationship differed based on candidate preference. Trump supporters' forecast happiness more strongly related to beliefs that their candidate would save society than for Biden supporters. Biden supporters' forecast anger and fear were more strongly related to beliefs that Trump would destroy society than vice versa. These forecasts mattered: predictions of lower happiness and greater anger if the non-preferred candidate won predicted voting, with Biden supporters voting more than Trump supporters. Generally, participants forecast more emotion than they experienced, but beliefs altered this tendency. Stronger beliefs predicted experiencing more happiness or more anger and fear about the election outcome than had been forecast. These findings have implications for understanding the mechanisms through which political polarisation and rhetoric can influence voting behaviour.

Reliable detection of generalized convulsive seizures using an off-the-shelf digital watch: A multisite phase 2 study.

OBJECTIVE: The aim of this study was to develop a machine learning algorithm using an off-the-shelf digital watch, the Samsung watch (SM-R800), and evaluate its effectiveness for the detection of generalized convulsive seizures (GCS) in persons with epilepsy. METHODS: This multisite epilepsy monitoring unit (EMU) phase 2 study included 36 adult patients. Each patient wore a Samsung watch that contained accelerometer, gyroscope, and photoplethysmographic sensors. Sixty-eight time and frequency domain features were extracted from the sensor data and were used to train a random forest algorithm. A testing framework was developed that would better reflect the EMU setting, consisting of (1) leave-one-patient-out cross-validation (LOPO CV) on GCS patients, (2) false alarm rate (FAR) testing on nonseizure patients, and (3) "fixed-and-frozen" prospective testing on a prospective patient cohort. Balanced accuracy, precision, sensitivity, and FAR were used to quantify the performance of the algorithm. Seizure onsets and offsets were determined by using video-electroencephalographic (EEG) monitoring. Feature importance was calculated as the mean decrease in Gini impurity during the LOPO CV testing. RESULTS: LOPO CV results showed balanced accuracy of .93 (95% confidence interval [CI] = .8-.98), precision of .68 (95% CI = .46-.85), sensitivity of .87 (95% CI = .62-.96), and FAR of .21/24 h (interquartile range [IQR] = 0-.90). Testing the algorithm on patients without seizure resulted in an FAR of .28/24 h (IQR = 0-.61). During the "fixed-and-frozen" prospective testing, two patients had three GCS, which were detected by the algorithm, while generating an FAR of .25/24 h (IQR = 0-.89). Feature importance showed that heart rate-based features outperformed accelerometer/gyroscope-based features. SIGNIFICANCE: Commercially available wearable digital watches that reliably detect GCS, with minimum false alarm rates, may overcome usage adoption and other limitations of custom-built devices. Contingent on the outcomes of a prospective phase 3 study, such devices have the potential to provide non-EEG-based seizure surveillance and forecasting in the clinical setting.

Application of extreme learning machine (ELM) forecasting model on CO2 emission dataset of a natural gas-fired power plant in Dhaka, Bangladesh.

Understanding and predicting CO2 emissions from individual power plants is crucial for developing effective mitigation strategies. This study analyzes and forecasts CO2 emissions from an engine-based natural gas-fired power plant in Dhaka Export Processing Zone (DEPZ), Bangladesh. This study also presents a rich dataset and ELM-based prediction model for a natural gas-fired plant in Bangladesh. Utilizing a rich dataset of Electricity generation and Gas Consumption, CO2 emissions in tons are estimated based on the measured energy use, and the ELM models were trained on CO2 emissions data from January 2015 to December 2022 and used to forecast CO2 emissions until December 2026. This study aims to improve the understanding and prediction of CO2 emissions from natural gas-fired power plants. While the specific operational strategy of the studied plant is not available, the provided data can serve as a valuable baseline or benchmark for comparison with similar facilities and the development of future research on optimizing operations and CO2 mitigation strategies. The Extreme Learning Machine (ELM) modeling method was employed due to its efficiency and accuracy in prediction. The ELM models achieved performance metrics Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Scaled Error (MASE), values respectively 3494.46 (<5000), 2013.42 (<2500), and 0.93 close to 1, which falls within the acceptable range. Although natural gas is a cleaner alternative, emission reduction remains essential. This data-driven approach using a Bangladeshi case study provides a replicable framework for optimizing plant operations and measuring and forecasting CO2 emissions from similar facilities, contributing to global climate change.

A Novel Deep Learning Approach for Forecasting Myocardial Infarction Occurrences with Time Series Patient Data.

Myocardial Infarction (MI) commonly referred to as a heart attack, results from the abrupt obstruction of blood supply to a section of the heart muscle, leading to the deterioration or death of the affected tissue due to a lack of oxygen. MI, poses a significant public health concern worldwide, particularly affecting the citizens of the Chittagong Metropolitan Area. The challenges lie in both prevention and treatment, as the emergence of MI has inflicted considerable suffering among residents. Early warning systems are crucial for managing epidemics promptly, especially given the escalating disease burden in older populations and the complexities of assessing present and future demands. The primary objective of this study is to forecast MI incidence early using a deep learning model, predicting the prevalence of heart attacks in patients. Our approach involves a novel dataset collected from daily heart attack incidence Time Series Patient Data spanning January 1, 2020, to December 31, 2021, in the Chittagong Metropolitan Area. Initially, we applied various advanced models, including Autoregressive Integrated Moving Average (ARIMA), Error-Trend-Seasonal (ETS), Trigonometric seasonality, Box-Cox transformation, ARMA errors, Trend and Seasonal (TBATS), and Long Short Time Memory (LSTM). To enhance prediction accuracy, we propose a novel Myocardial Sequence Classification (MSC)-LSTM method tailored to forecast heart attack occurrences in patients using the newly collected data from the Chittagong Metropolitan Area. Comprehensive results comparisons reveal that the novel MSC-LSTM model outperforms other applied models in terms of performance, achieving a minimum Mean Percentage Error (MPE) score of 1.6477. This research aids in predicting the likely future course of heart attack occurrences, facilitating the development of thorough plans for future preventive measures. The forecasting of MI occurrences contributes to effective resource allocation, capacity planning, policy creation, budgeting, public awareness, research identification, quality improvement, and disaster preparedness.

Explaining empirical dynamic modelling using verbal, graphical and mathematical approaches.

Empirical dynamic modelling (EDM) is becoming an increasingly popular method for understanding the dynamics of ecosystems. It has been applied to laboratory, terrestrial, freshwater and marine systems, used to forecast natural populations and has addressed fundamental ecological questions. Despite its increasing use, we have not found full explanations of EDM in the ecological literature, limiting understanding and reproducibility. Here we expand upon existing work by providing a detailed introduction to EDM. We use three progressively more complex approaches. A short verbal explanation of EDM is then explicitly demonstrated by graphically working through a simple example. We then introduce a full mathematical description of the steps involved. Conceptually, EDM translates a time series of data into a path through a multi-dimensional space, whose axes are lagged values of the time series. A time step is chosen from which to make a prediction. The state of the system at that time step corresponds to a 'focal point' in the multi-dimensional space. The set (called the library) of candidate nearest neighbours to the focal point is constructed, to determine the nearest neighbours that are then used to make the prediction. Our mathematical explanation explicitly documents which points in the multi-dimensional space should not be considered as focal points. We suggest a new option for excluding points from the library that may be useful for short-term time series that are often found in ecology. We focus on the core simplex and S-map algorithms of EDM. Our new R package, pbsEDM, enhances understanding (by outputting intermediate calculations), reproduces our results and can be applied to new data. Our work improves the clarity of the inner workings of EDM, a prerequisite for EDM to reach its full potential in ecology and have wide uptake in the provision of advice to managers of natural resources.

Multi-step forecasting of dissolved oxygen in River Ganga based on CEEMDAN-AdaBoost-BiLSTM-LSTM model.

Accurate prediction of Dissolved Oxygen (DO) is an integral part of water resource management. This study proposes a novel approach combining Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) with AdaBoost and deep learning for multi-step forecasting of DO. CEEMDAN generates Intrinsic Mode Functions (IMFs) with different frequencies, capturing non-linear and non-stationary characteristics of the data. The high-frequency and medium-frequency IMFs, characterized by complex patterns and frequent changes over time, are predicted using Adaboost with Bidirectional Long Short-Term Memory (BiLSTM) as the base estimator. The low-frequency IMFs, characterized by relatively simple patterns, are predicted using standalone Long Short-Term Memory (LSTM). The proposed CEEMDAN-AdaBoost-BiLSTM-LSTM model is tested on data from ten stations of river Ganga. We compare the results with six models without decomposition and four models utilizing decomposition. Experimental results show that using a tailored prediction technique based on each IMF's distinctive features leads to more accurate forecasts. CEEMDAN-AdaBoost-BiLSTM-LSTM outperforms CEEMDAN-BiLSTM with an average improvement of 25.458% for RMSE and 37.390% for MAE. Compared with CEEMDAN-AdaBoost-BiLSTM, an average improvement of 20.779% for RMSE and 28.921% for MAE is observed. Diebold-Mariano test and t-test suggest a statistically significant difference in performance between the proposed and compared models.

Exploration on human resource management and prediction model of data-driven information security in Internet of Things.

The advent of the Internet of Things (IoT) has accelerated the pace of economic development across all sectors. However, it has also brought significant challenges to traditional human resource management, revealing an increasing number of problems and making it unable to meet the needs of contemporary enterprise management. The IoT has brought numerous conveniences to human society, but it has also led to security issues in communication networks. To ensure the security of these networks, it is necessary to integrate data-driven technologies to address this issue. In response to the current state of human resource management, this paper proposes the application of IoT technology in enterprise human resource management and combines it with radial basis function neural networks to construct a model for predicting enterprise human resource needs. The model was also experimentally analyzed. The results show that under this algorithm, the average prediction accuracy for the number of employees over five years is 90.2 %, and the average prediction accuracy for sales revenue is 93.9 %. These data indicate that the prediction accuracy of the model under this study's algorithm has significantly improved. This paper also conducted evaluation experiments on a wireless communication network security risk prediction model. The average prediction accuracy of four tests is 91.21 %, indicating that the model has high prediction accuracy. By introducing data-driven technology and IoT applications, this study provides new solutions for human resource management and communication network security, promoting technological innovation in the fields of traditional human resource management and information security management. The research not only improves the accuracy of the prediction models but also provides strong support for decision-making and risk management in related fields, demonstrating the great potential of big data and artificial intelligence technology in the future of enterprise management and security.

Demand-side load forecasting in smart grids using machine learning techniques.

Electrical load forecasting remains an ongoing challenge due to various factors, such as temperature and weather, which change day by day. In this age of Big Data, efficient handling of data and obtaining valuable information from raw data is crucial. Through the use of IoT devices and smart meters, we can capture data efficiently, whereas traditional methods may struggle with data management. The proposed solution consists of two levels for forecasting. The selected subsets of data are first fed into the "Daily Consumption Electrical Networks" (DCEN) network, which provides valid input to the "Intra Load Forecasting Networks" (ILFN) network. To address overfitting issues, we use classic or conventional neural networks. This research employs a three-tier architecture, which includes the cloud layer, fog layer, and edge servers. The classical state-of-the-art prediction schemes usually employ a two-tier architecture with classical models, which can result in low learning precision and overfitting issues. The proposed approach uses more weather features that were not previously utilized to predict the load. In this study, numerous experiments were conducted and found that support vector regression outperformed other methods. The results obtained were 5.055 for mean absolute percentage error (MAPE), 0.69 for root mean square error (RMSE), 0.37 for normalized mean square error (NRMSE), 0.0072 for mean squared logarithmic error (MSLE), and 0.86 for R2 score values. The experimental findings demonstrate the effectiveness of the proposed method.

Dynamic hierarchical state space forecasting.

In this paper, we aim to both borrow information from existing units and incorporate the target unit's history data in time series forecasting. We consider a situation when we have time series data from multiple units that share similar patterns when aligned in terms of an internal time. The internal time is defined as an index according to evolving features of interest. When mapped back to the calendar time, these time series can span different time intervals that can include the future calendar time of the targeted unit, over which we can borrow the information from other units in forecasting the targeted unit. We first build a hierarchical state space model for the multiple time series data in terms of the internal time, where the shared components capture the similarities among different units while allowing for unit-specific deviations. A conditional state space model is then constructed to incorporate the information of existing units as the prior information in forecasting the targeted unit. By running the Kalman filtering based on the conditional state space model on the targeted unit, we incorporate both the information from the other units and the history of the targeted unit. The forecasts are then transformed from internal time back into calendar time for ease of interpretation. A simulation study is conducted to evaluate the finite sample performance. Forecasting state-level new COVID-19 cases in United States is used for illustration.

Development, Validation and Clinical Utility of a Risk Prediction Model for Maternal and Neonatal Adverse Outcomes in Pregnant Women with Hypothyroidism.

Purpose: This study aimed to create, verify and assess the clinical utility of a prediction model for maternal and neonatal adverse outcomes in pregnant women with hypothyroidism. Methods: A prediction model was developed, and its accuracy was tested using data from a retrospective cohort. The study focused exclusively on female patients diagnosed with hypothyroidism who were admitted to a tertiary hospital. The development and validation cohort comprised individuals who gave birth between 1 October 2020 and 31 December 2022. The primary outcome was a combination of crucial maternal and newborn problems (eg premature births, abortions and neonatal asphyxia). The prediction model was developed using logistic regression. Evaluation of the model's performance was conducted based on its ability to discriminate, calibrate and provide clinical value. Results: In total, nine variables were chosen to develop the predictive model for adverse maternal and neonatal outcomes during pregnancy with hypothyroidism. The area under the curve of the model for predicting maternal adverse outcomes was 0.845, and that for predicting neonatal adverse outcomes was 0.685. The calibration plots showed good agreement between the nomogram predictions and the actual observations in both the training and validation cohorts. Furthermore, decision curve analysis suggested that the nomograms were clinically useful and had good discriminative power to identify high-risk mother-infant cases. Conclusion: Two models to predict the risk probability of maternal and neonatal adverse outcomes in pregnant women with hypothyroidism were developed and verified to assist physicians in evaluating maternal and neonatal adverse outcomes throughout pregnancy with hypothyroidism and to facilitate decision-making regarding therapy.

Application of a hybrid algorithm of LSTM and Transformer based on random search optimization for improving rainfall-runoff simulation.

Flood forecasting using traditional physical hydrology models requires consideration of multiple complex physical processes including the spatio-temporal distribution of rainfall, the spatial heterogeneity of watershed sub-surface characteristics, and runoff generation and routing behaviours. Data-driven models offer novel solutions to these challenges, though they are hindered by difficulties in hyperparameter selection and a decline in prediction stability as the lead time extends. This study introduces a hybrid model, the RS-LSTM-Transformer, which combines Random Search (RS), Long Short-Term Memory networks (LSTM), and the Transformer architecture. Applied to the typical Jingle watershed in the middle reaches of the Yellow River, this model utilises rainfall and runoff data from basin sites to simulate flood processes, and its outcomes are compared against those from RS-LSTM, RS-Transformer, RS-BP, and RS-MLP models. It was evaluated against RS-LSTM, RS-Transformer, RS-BP, and RS-MLP models using the Nash-Sutcliffe Efficiency Coefficient (NSE), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Bias percentage as metrics. At a 1-h lead time during calibration and validation, the RS-LSTM-Transformer model achieved NSE, RMSE, MAE, and Bias values of 0.970, 14.001m3/s, 5.304m3/s, 0.501% and 0.953, 14.124m3/s, 6.365m3/s, 0.523%, respectively. These results demonstrate the model's superior simulation capabilities and robustness, providing more accurate peak flow forecasts as the lead time increases. The study highlights the RS-LSTM-Transformer model's potential in flood forecasting and the advantages of integrating various data-driven approaches for innovative modelling.

Local spatial and temporal relation discovery model based on attention mechanism for traffic forecasting.

Recognizing the evolution pattern of traffic condition and making accurate prediction play a vital role in intelligent transportation systems (ITS). With the massive increase of available traffic data, deep learning-based models have attracted considerable attention for their impressive performance in traffic forecasting. However, the majority of existing approaches neglect to model of asynchronously dynamic spatio-temporal correlation and fail to consider the impact of historical traffic data on future condition. Additionally, the attribute of deep learning method presents challenges in interpreting the explicit spatiotemporal relationships. In order to enhance the accuracy of traffic prediction as well as extract comprehensive and explainable spatial-temporal relevance in traffic networks, we propose a novel attention-based local spatial and temporal relation discovery (ALSTRD) model. Our model firstly implements feature representation learning to effectively express latent input traffic information. Then, a local attention mechanism structure is established to model asynchronous dependencies of historical input data. Finally, another attention network and the Pearson Correlation Coefficient method are introduced to extract the elaborate influence of the historical traffic condition of neighboring roads on the future condition of the target road. The experiment results on several datasets demonstrate that our model achieves significant improvements in prediction accuracy compared to other baseline methods, which can be attributed to its ability to extract the fine-grained correlation among historical traffic data and capture the dynamic association between past and future data. In addition, the incorporation of attention mechanism and Pearson Correlation Coefficient promotes the model's ability to elucidate spatiotemporal correlations among traffic data, thereby providing a more robust explanation.

A novel seasonal grey prediction model with fractional order accumulation for energy forecasting.

To accurately predict sequence data with seasonal characteristics, we combine data restart technology and fractional order accumulation into a novel seasonal grey model (FSGM (1,1, α)). The particle swarm optimization (PSO) algorithm is used to solve the fractional order and background value coefficients of the model, and the effectiveness of FSGM (1,1, α) is verified using three cases. Finally, we use FSGM (1,1, α) to predict quarterly electricity generation in Beijing and Henan Province and quarterly petroleum coke production in China from 2023 to 2027. The research results indicate that, first, FSGM (1,1, α) is reasonable and effective and has the ability to accurately capture the dynamic trend of seasonal data. Second, compared with the grey model (GM (1,1)), seasonal grey model (SGM (1,1)), data grouping grey model (DGGM (1,1)), data grouping seasonal model (DGSM (1,1)), and data grouping seasonal time model (DGSTM (1,1)), which have seasonal characteristics, FSGM (1,1, α) can better fit the original data, achieve higher prediction accuracy, and perform better. Third, from 2023 to 2027, it is predicted that there will be no significant change in Beijing's electricity generation, and the current stable trend will be maintained. Both the power generation in Henan Province and the petroleum coke production in China will steadily increase to a certain extent, with obvious seasonal cyclical fluctuations. Notably, the power generation and petroleum coke production in Henan Province in the fourth quarter of 2027 will increase by 11.50 % and 10.93 %, respectively, compared to those in the fourth quarter of 2023.