Time Series Regression Models Research Articles

Dental Pain Medication Prescriptions in Minas Gerais, Brazil (2011-2021): A Time-Series Analysis.

To describe trends of dentist-prescribed non-steroidal anti-inflammatory drugs (NSAIDs) and analgesics, from January 2011 to December 2021, as well as to examine the relationship between these trends and characteristics of public oral health services in Minas Gerais, Brazil. In this time-series analysis, all drugs were classified according to the Anatomical Therapeutic Chemical classification system. Drugs categorized as NSAIDs (M01A), and other analgesics and antipyretics (N02B) were included for analysis. The outcome was the number of Defined Daily Doses (DDDs)/1000 inhabitants/year for NSAIDs and analgesics in each town. Covariates referred to characteristics of public oral health services, such as coverage, estimates of dental procedures, and frequency of toothache. Linear time-series regression models were used to determine the influence of covariates on the outcome. Overall, there were 58,482 prescriptions of NSAIDs recorded in thirty-eight towns, while 47,499 prescriptions of analgesics in forty-three towns. For each year, there was a 0.38 (p < 0.001), and 0.28 (p < 0.001) increase in the average log of DDD/1000 inhabitants/year for NSAIDs and analgesics, respectively. A positive association was detected between toothache (p < 0.001) and the prescription of NSAIDs. Over the eleven years, there was a general rising trend in the prescriptions. Toothache was the only characteristic of public oral health services associated with the prescription rates of NSAIDs, implying that as the frequency of toothaches increase, so do the prescriptions of NSAIDs in the studied towns.

Heat exposure and hospitalizations for chronic kidney disease in China: a nationwide time series study in 261 major Chinese cities

BackgroundClimate change profoundly shapes the population health at the global scale. However, there was still insufficient and inconsistent evidence for the association between heat exposure and chronic kidney disease (CKD).MethodsIn the present study, we studied the association of heat exposure with hospitalizations for cause-specific CKD using a national inpatient database in China during the study period of hot season from 2015 to 2018. Standard time-series regression models and random-effects meta-analysis were developed to estimate the city-specific and national averaged associations at a 7 lag-day span, respectively.ResultsA total of 768,129 hospitalizations for CKD was recorded during the study period. The results showed that higher temperature was associated with elevated risk of hospitalizations for CKD, especially in sub-tropical cities. With a 1 °C increase in daily mean temperature, the cumulative relative risks (RR) over lag 0–7 d were 1.008 [95% confidence interval (CI) 1.003–1.012] for nationwide. The attributable fraction of CKD hospitalizations due to high temperatures was 5.50%. Stronger associations were observed among younger patients and those with obstructive nephropathy. Our study also found that exposure to heatwaves was associated with added risk of hospitalizations for CKD compared to non-heatwave days (RR = 1.116, 95% CI 1.069–1.166) above the effect of daily mean temperature.ConclusionsShort-term heat exposure may increase the risk of hospitalization for CKD. Our findings provide insights into the health effects of climate change and suggest the necessity of guided protection strategies against the adverse effects of high temperatures.

Time Series Regression: Prediction of Electricity Consumption Based on Number of Consumers at National Electricity Supply Company

Electrical energy is one of the components of Gross Domestic Product that is able to encourage the economy because it has become a basic need of the community. To meet the increasing demand for electrical energy, the Indonesia National Electricity Providers (PLN) need to predict the amount of electrical power required based on the customer numbers to meet the demand for adequate electricity supply. This study aims to predict electric power based on electricity user customers using a time series regression model. The data used in this study are secondary data which get from PLN annual report in 2021. This study resulted in a finding of the best prediction model based on the Akaike Information Criterion (AIC) value, namely the time series regression model with the error value modeled by the AR(1) model, while the forecasting accuracy measure used the value MAPE of 9.77%. This means that the result of model prediction is highly accurate.

Time Series Regression Modeling with AR(1) Errors

When ordinary regression analysis is performed using time-series variables, it is common for the errors (residuals) to have a time-series structure. This violates the usual assumption of independent errors in ordinary least squares (OLS) regressions. Consequently, the estimates of the coefficients and their standard errors are incorrect if the time-series structure of the errors is ignored. In this study, an investigation of a regression model with time-series variables, particularly a simple case, was conducted using the conventional method. The ‘AirPassengers Dataset’ was downloaded from the R repository used for the analysis. Ordinary least squares and Cochrane-Orcutt procedures were used as methodologies. The results show that the adjusted regression model with autoregressive errors outperformed the ordinary regression model.

Trends and Prescriber Variation in the Duration of Oxaliplatin-Containing Adjuvant Chemotherapy for Stage III Colon Cancer From 2007 to 2019: A Population-Based Retrospective Cohort Study

The International Duration Evaluation of Adjuvant Therapy (IDEA) collaboration in 2017 established 3 months of adjuvant therapy as an alternative to 6 months of therapy for stage III colon cancer. We determined the association between the IDEA publication, changes in clinical practice, and prescriber variation. Using linked databases, we identified Ontarians aged ≥18 years at diagnosis of stage III colon cancer between 2007 and 2019 who received oxaliplatin-containing adjuvant therapy. The outcome was duration of therapy, categorized as ≤25%, >25% to ≤50%, >50% to ≤75%, and >75% of a 6-month course of therapy to approximate treatment durations in the IDEA collaboration. We examined trends in duration over time using an interrupted time series regression model. We analyzed treatment duration after accounting for patient and prescriber characteristics, using multivariable mixed effects logistic regression models to quantify between-prescriber variation. We included 4695 patients with stage III colon cancer who received oxaliplatin-containing adjuvant chemotherapy, of whom 77.5% initiated treatment pre-IDEA and 22.5% initiated treatment post-IDEA. Post-IDEA, there was a 16.4% (95% CI, 12.5%-20.3%) absolute increase in the proportion of patients treated with ≤50% of a maximal course of therapy. This trend was greatest among patients with low-risk tumors. Prescriber variation increased pre-IDEA to 15.6% post-IDEA (variance partition coefficient 5.4% pre-IDEA and 15.6% post-IDEA). The publication of IDEA was associated with increases in short duration adjuvant therapy and prescriber-level practice variation for stage III colon cancer. Clinicians should be better supported to make consistent recommendations about adjuvant duration under conditions of uncertainty and trade-offs.

Analysis of Foreign Direct Investment Net Inflows in the Philippines: Trends, Predictors, and Forecasting Using Time Series and Linear Regression Models

Analysis of Foreign Direct Investment Net Inflows in the Philippines: Trends, Predictors, and Forecasting Using Time Series and Linear Regression Models

Uncertainty quantification of a deep learning model for failure rate prediction of water distribution networks

Predicting the time-dependent pipe failure rate of the water distribution networks (WDNs) is important for planning its renewal budget but also challenging due to the complex factors involved. The recent development of machine learning techniques provides a novel approach for accurate failure rate prediction based on historical data. However, the inherited randomness and uncertainty of water pipe failures and machine learning algorithms are often ignored in the training and prediction process. This article develops an uncertainty quantification framework for deep learning-based WDN system-wide failure rate prediction. The framework integrates a probabilistic long short-term memory (LSTM) model with a Monte Carlo method. The historical climate data and WDN pipe maintenance data over the past 35 years for Cuyahoga County, USA, are used to illustrate this deep-learning model-based uncertainty quantification framework in this study. A statistical time series regression model, ARIMAX, is used as a comparison benchmark. The results show that the LSTM model outperforms the ARIMAX model in prediction accuracy in most years by considering the uncertainties. Besides, the uncertainty range of the LSTM model prediction is 50% of that of the ARIMAX model. The results also identified the major contributing factors to the uncertainties of LSTM machine learning model prediction. The proposed uncertainty framework features excellent extensibility and can be adapted to quantify uncertainties with other types of machine learning models.

Reconstruction of Missing Well-Logs Using Facies-Informed Discrete Wavelet Transform and Time Series Regression

Summary Geophysical logging is widely used in lithofacies identification, reservoir parameter prediction, and geological modeling. However, it is common to have well-log sections with low-quality and/or missing segments. Repeating the well-log measurements is not only expensive but might also be impossible depending on the condition of the borehole walls. In these situations, reliable and accurate well-log prediction is, therefore, necessary in different stages of the geomodeling workflow. In this study, we propose a time series regression model to predict missing well-log data, incorporating facies information as an additional geological input and using discrete wavelet transform (DWT) to denoise the input data set. The main contributions of this work are threefold: (i) We jointly use facies information with well logs as the input data set; (ii) we use DWT to denoise the input data and consequently improve the signal-to-noise ratio of the input data; and (iii) we regard the depth domain as the time domain and use a time series regression algorithm for log reconstruction modeling. We show a real application example in two distinct scenarios. In the first, we predict missing well-log intervals. In the second, we predict complete well logs. The experimental results show the ability of the proposed prediction model to recover missing well-log data with high accuracy levels.

Promoting physical activity within breast cancer survivors using behavior change techniques: N-of-1 randomized controlled trials.

Survivors of breast cancer are an important target for health promotion interventions, and physical activity (PA) is recommended with benefits being well established. More and more behavior theory guided interventions are planned and tailored to the individuals, targeting specifics needs and personal characteristics. We aimed to test the effectiveness of two behavior change techniques [coping planning (CP) and self-monitoring (SM)] for promoting PA in survivors of breast cancer, using a single-case design. In a powerful multiple sample N-of-1 RCT, 10 women post-breast cancer (M = 52.8 years) were randomly assigned to four experimental conditions (CP and/or SM, control) for 60 days and their daily step count was observed. The overall effects were analyzed using multilevel time-series with random intercepts. Time-series regression models and supplementary pairwise analyses were conducted for individuals. Multilevel analyses showed significant effects of CP and SM. Single-case analyses showed that six participants walked significantly more on the intervention days, that combined interventions were effective for five participants, but two participants had better results with the other two conditions. Combined self-regulatory techniques were more effective in promoting PA however, individual variability should be considered in future studies with survivors of breast cancer. Retrospectively registered: https://osf.io/eakjy/?view_only=81bf3dc084a0410a8ea28265e0f3c360, 31-01-2023.(raise query).

Cause-specific mortality and burden attributable to temperature variability in China

BackgroundFew large-scale, nationwide studies have assessed cause-specific mortality risks and burdens associated with temperature variability (TV). ObjectiveTo estimate associations between TV and cause-specific mortality and quantify the mortality burden in China. MethodsData on daily total and cause-specific mortality in 272 Chinese cities between 2013 and 2015 were recorded. TVs were computed as the standard deviations of daily minimum and maximum temperatures over a duration of 2 to 7 days. The time-series quasi-Poisson regression model with adjustment of the cumulative effects of daily mean temperature over the same duration was applied to evaluate the city-specific associations of TV and mortality. Then, we pooled the effect estimates using a random-effects meta-analysis and calculated the mortality burdens. ResultsOverall, TV showed significant and positive associations with total and cause-specific mortality. The TV-mortality associations were generally stronger when using longer durations. A 1 °C increase in TV at 0–7 days (TV0–7) was associated with a 0.79 % [95 % confidence interval (CI): 0.55 %, 0.96 %] increase in total mortality. Mortality fractions attributable to TV0–7 were 4.37 % for total causes, 4.75 % for overall cardiovascular disease, 4.37 % for coronary heart disease, 5.05 % for stroke, 8.28 % for ischaemic stroke, 1.08 % for haemorrhagic stroke, 6.93 % for respiratory disease, and 6.81 % for COPD, respectively. The mortality risk and burden were generally higher in the temperate monsoon zone, females, and elders. ConclusionThis nationwide study indicated that TV was an independent risk factor of mortality, and could result in significant burden for main cardiorespiratory diseases.

A novel framework for machine intelligence aided smarter energy management in residential scenario

Energy management systems can monitor, optimize, and control energy utilization in residential and commercial applications. Different solutions are available for achieving better energy management. Among the different solutions, load forecasting is considered as one of most preferred technique. It is essential for a load forecasting technique to predict the aggregated loads at the residential level. This paper presents a systematic analysis of different load forecasting models considering a hybrid renewable energy based smart home which uses a solar photovoltaic (PV) system, wind system, and battery storage system for energy storage. In this work, the various forecasting techniques such as time-series models, regression models, ARIMA (Autoregressive Integrated Moving Average) models, and neural network models are analyzed for different load cases. Simulation is performed based on the effect of seasonal bias and condition load shifting on energy generation. The obtained results show that both polynomial and SVM regression model achieves better performance in terms of R2 score and adjusted R2 score of 0.73, and RMSE values of 0.84. It can also be inferred from the analysis that, the random forest regression model and a neural network-based-long short-term memory (LSTM) model exhibit superior results in comparison to other models.

The indirect impacts of nonpharmacological COVID-19 control measures on other infectious diseases in Yinchuan, Northwest China: a time series study

BackgroundVarious nonpharmaceutical interventions (NPIs) against COVID-19 continue to have an impact on socioeconomic and population behaviour patterns. However, the effect of NPIs on notifiable infectious diseases remains inconclusive due to the variability of the disease spectrum, high-incidence endemic diseases and environmental factors across different geographical regions. Thus, it is of public health interest to explore the influence of NPIs on notifiable infectious diseases in Yinchuan, Northwest China.MethodsBased on data on notifiable infectious diseases (NIDs), air pollutants, meteorological data, and the number of health institutional personnel in Yinchuan, we first fitted dynamic regression time series models to the incidence of NIDs from 2013 to 2019 and then estimated the incidence for 2020. Then, we compared the projected time series data with the observed incidence of NIDs in 2020. We calculated the relative reduction in NIDs at different emergency response levels in 2020 to identify the impacts of NIPs on NIDs in Yinchuan.ResultsA total of 15,711 cases of NIDs were reported in Yinchuan in 2020, which was 42.59% lower than the average annual number of cases from 2013 to 2019. Natural focal diseases and vector-borne infectious diseases showed an increasing trend, as the observed incidence in 2020 was 46.86% higher than the estimated cases. The observed number of cases changed in respiratory infectious diseases, intestinal infectious diseases and sexually transmitted or bloodborne diseases were 65.27%, 58.45% and 35.01% higher than the expected number, respectively. The NIDs with the highest reductions in each subgroup were hand, foot, and mouth disease (5854 cases), infectious diarrhoea (2157 cases) and scarlet fever (832 cases), respectively. In addition, it was also found that the expected relative reduction in NIDs in 2020 showed a decline across different emergency response levels, as the relative reduction dropped from 65.65% (95% CI: -65.86%, 80.84%) during the level 1 response to 52.72% (95% CI: 20.84%, 66.30%) during the level 3 response.ConclusionsThe widespread implementation of NPIs in 2020 may have had significant inhibitory effects on the incidence of respiratory infectious diseases, intestinal infectious diseases and sexually transmitted or bloodborne diseases. The relative reduction in NIDs during different emergency response levels in 2020 showed a declining trend as the response level changed from level 1 to level 3. These results can serve as essential guidance for policy-makers and stakeholders to take specific actions to control infectious diseases and protect vulnerable populations in the future.

Hybrid ARDL-MIDAS-Transformer time-series regressions for multi-topic crypto market sentiment driven by price and technology factors

This paper develops a novel hybrid Autoregressive Distributed Lag Mixed Data Sampling (ARDL-MIDAS) model that integrates both deep neural network multi-head attention Transformer mechanisms, and a number of covariates, including sophisticated stochastic text time-series features, into a mixed-frequency time-series regression model with long memory structure. In doing so, we demonstrate how the resulting class of ARDL-MIDAS-Transformer models allows one to maintain the interpretability of the time-series models whilst exploiting the deep neural network attention architectures. The latter may be used for higher-order interaction analysis, or, as in our use case, for design of Instrumental Variables to reduce bias in the estimation of the infinite lag ARDL-MIDAS model. Our approach produces an accurate, interpretable forecasting framework that allows one to forecast end-of-day sentiment intra-daily, with readily attainable time-series regressors. In this regard, we conduct a statistical time-series analysis on mixed data frequencies to discover and study the relationships between sentiment from our custom stochastic text time-series sentiment framework, alternative popular sentiment extraction frameworks (BERT and VADER), and technology factors, as well as to investigate the role that price discovery has on retail cryptocurrency investors’ sentiment (crypto sentiment). This is an interesting time-series modelling challenge as it involves working with time-series regression models in which the time-series response process, and the regression time-series covariates, are observed at different time scales. Specifically, a detailed real-data study is conducted where we explore the relationship between daily crypto market sentiment (of positive, negative and neutral polarity) and the intra-daily (hourly) price log-return dynamics of crypto markets. The sentiment indices constructed for a variety of “topics” and news sources are produced as a collection of time-series capturing the daily sentiment polarity signals for each “topic”, namely each particular market or crypto asset. Different sentiment methods are developed in a time-series context, and utilised in the proposed hybrid regression framework. Furthermore, technology factors are introduced to capture network effects, such as the hash rate which is an important aspect of the money supply relating to the mining of new crypto assets, and block hashing for transaction verification. Throughout our real data study, we provide guidance and insights on how to use our hybrid model to combine—in a transparent, non-black-box way—covariates obtained with different time resolutions, how to understand the arising dynamics between these covariates, potentially under the presence of long memory structure, and, finally, successfully leverage these in forecasting applications. The hybrid model developed demonstrated superior performance to alternatives in both in-sample and forecasting application on real data.

Association between airborne particulate matter（PM 2.5） concentration and the incidence of allergic rhinitis in Shanghai

Objective:To explore the impact of PM 2.5 concentration in Shanghai on the incidence of allergic rhinitis（AR） in the population, and provide strategies for early warning and prevention of AR. Methods:Collect daily average concentrations of atmospheric pollutants monitored in Shanghai from January 1, 2017 to December 31, 2019, and clinical data of AR patients from five hospitals in Shanghai during the same period. We used a time-series analysis additive Poisson regression model to analyze the correlation between PM 2.5 levels and outpatient attendance for AR patients. Results:During the study period, a total of 56 500 AR patients were included, and the daily average concentration of PM 2.5 was（35.28±23.07）μg/m³. There is a correlation between the concentration of PM 2.5 and the number of outpatient attendance for AR cases. There is a positive correlation between the daily average number of outpatient for AR and levels of PM 2.5 air pollution（（P<0.05）） . We found that every 10 μg/m³ increase in PM 2.5, the impact of on the number of AR visits was statistically significant on the same day, the first day behind, and the second day behind, with the strongest impact being the exposure on the same day. Every 10 μg/m³ increases in PM 2.5, the number of outpatient visits increased by 0.526% on the same day（95%CI 1.000 50-1.010 04）. Conclusion:The atmospheric PM 2.5 concentration in Shanghai is positively correlated with the number of outpatient for AR, and PM 2.5 exposure is an independent factor in the onset of AR. This provides an important theoretical basis for AR.

Spatial Variability of Summer Temperature and Related All-cause Mortality from 2006 to 2015 for Indian Cities: A Time Series Analysis

Abundant literature is available on an extremely high temperature associated with mortality for cities of the developed world, but there is a dearth in the literature for coastal, desert and dry cities of the developing world, especially for India. We examined all-cause mortality and extreme high temperature in three Indian cities representing coastal, desert and dry areas for summer months (March to June) from 2006 to 2015. We obtained the data on temperature and all-cause mortality for ten years for the summer months. The city-specific effect of ambient heat on all-cause mortality was assessed through time series ordinary least square linear regression model. A total of 75,571, 122,117 and 53,042 deaths for 1,203, 1,220 and 1,180 summer days from 2006 to 2015 were analysed with ambient temperature for Jaipur, Hyderabad and Surat, respectively. There were 994 (27.6%) out of 3,603 summer days having temperature ≥40°C and 2,495 (69.3%) out of 3,602 summer days having feel temperature/heat index (HI) of ≥41°C. According to the Indian Meteorological Department (IMD) criteria for the heatwave, Surat has the maximum number of 75 days with a maximum temperature of ≥40°C, whereas Hyderabad has only 4 days and Jaipur faced 35 days with a maximum temperature of ≥45°C during the study period. The per-day mean all-cause mortality increased to 39% and 11% for Jaipur and Hyderabad, respectively, at ≥45°C and 20% for the coastal city of Surat at ≥40°C as per IMD heatwave criteria. A time-series linear regression model shows that adjusted R-squared is 0.593, 0.629 and 0.348, which explained the variation of 59.3%, 62.9% and 34.8% for all-cause mortality (dependent variable) by independent variables (maximum temperature, humidity and HI) for Jaipur, Hyderabad and Surat, respectively. The maximum temperature threshold (cut-off) for all-cause mortality for Jaipur, Hyderabad and Surat is 42°C, 41°C and 40°C, respectively. The impact of ambient heat in the rise of all-cause mortality for all study sites was evident. Hence, findings support the efforts for reducing the public health burden of high ambient temperature through developing and implementing city-specific heat action plans.

Impacts of Atmospheric and Load Conditions on the Power Substation Equipment Temperature Model

Infrared thermography is a predictive maintenance tool used in substations to identify a disturbance in electrical equipment that could lead to poor operation and potential failure in the future. According to Joule’s law, the temperature of electrical equipment is proportional to the current flowing through it. Other external factors, such as solar incidence, air humidity, wind speed, and air temperature, can interfere with its operating temperatures. Based on this premise, this article aims to analyze the influence of atmospheric and load conditions on the operational cycle of thermography-monitored equipment in order to describe the operating temperature of the object using only external data and to show the impacts of external influences on the final temperature reached by the object. Five multivariate time series regression models were developed to describe the maximum equipment temperature. The final model achieved the best fit between the measured and model temperature based on the Akaike information criterion (AIC) metric, where all external variables were used to compose the model. The proposed model shows the impacts of each external factor on equipment temperature and could be used to create a predictive maintenance strategy for power substations to avoid failure.

Gender differences on healthcare accessibility and outcomes of a electronic inter-clinician consultation program at the cardiology department in a Galician Health Area.

To assess the longer-term results (hospital admissions and mortality) in women versus men referred to a cardiology department from primary care using an e-consultation in our outpatient care program. We selected 61,306 patients (30,312 women and 30,994 men) who visited the cardiology service at least once between 2010 and 2021: 69,1% (19,997 women and 20,462 men) were attended in e-consultation (from 2013 to 2021) and 30,9% (8,920 women and 9,136 men) in in-person consultations (from 2010 to 2012) without gender differences in the proportion of patients attended in each period. Using an interrupted time series regression model, we analysed the impact of incorporating e-consultation into the health care model and evaluated the elapsed time to cardiology care, heart failure (HF), cardiovascular (CV), and all-cause hospital admissions and mortality during the one-year after cardiology consultation. The introduction of e-consultation substantially decreased waiting times to cardiology care; during the in-person consultation period, the mean delay for cardiology care was 57.9 (24.8) days in men and 55.8 (22.8) days in women. During the e-consultation period, the waiting time to cardiology care was markedly reduced to 9.41 (4.02) days in men and 9.46 (4.18) in women. After e-consultation implantation, there was a significant reduction in the 1-year rate of hospital admissions and mortality, both in women and men iRR [IC95%]: 0.95 [0.93-0.96] for HF, 0.90 [0.89-0.91] for CV, and 0.70 [0.69-0.71] for all-cause hospitalization; and 0.93 [0.92-0.95] for HF, 0.86 [0.86-0.87] for CV, and 0.88 [0.87-0.89] for all-cause mortality in women; and 0.91 [0.89-0.92] for HF, 0.90 [0.89-0.91] for CV and 0.72 [0.71-0.73] for all-cause hospitalization; and 0.96 [0.93-0.97] for HF, 0.87 [95%CI: 0.86-0.87] for CV, and 0.87 [0.86-0.87] for all-cause mortality, in men. Compared with the in-person consultation period, an outpatient care program that includes an e-consultation significantly reduced waiting time to cardiology care and was safe, with a lower rate of hospital admissions and mortality in the first year, without significative gender differences.

Forecasting The Search Trends of The Keyword “Sarung Wadimor” In Indonesia on Google Trends Data Using Time Series Regression with Calender Variation and Arima Box-Jenkins

The impact of this 4.0 era is that data is growing and can be collected very easily and then reprocessed to obtain information. One of the search engines for various data and information that is often used is Google, causing a high search intensity and will further impact on increasing the amount of data generated by search engines. Google Trends is one of the official websites from Google that reflects or takes pictures of events in society based on search keywords. The search keyword that will be studied in this article is “Sarung Wadimor”. Therefore, the purpose of this research is to forecast the search trend for the keyword "Sarung Wadimor" which is interesting because the resulting time series data pattern shows a recurring pattern due to the effect of calendar variations which are thought to be related to the month of Ramadan. Forecasting modeling uses Autoregressive Integrated Moving Average (ARIMA) and Time Series Regression (TSR). The goodness of the model used in this article is the Mean Square Error (MSE), Root Mean Square Error (RMSE), and Symmetric Mean Absolute Percentage Error (SMAPE). Based on the results of the analysis, using three goodness-of-fit measures shows that the TSR model with the Calendar Variation of Ramadan + Month Periods has smaller MSE, RMSE, and SMAPE values than the other models with goodness-of-fit values of 88.602, 9.413, and 26.950, respectively. Forecasting results for the next 6 periods show that the search trend for the keyword "Sarung Wadimor" tends to decrease, this is because the month of Ramadan is still quite far in 2023.

On the Soundness of XAI in Prognostics and Health Management (PHM)

The aim of predictive maintenance, within the field of prognostics and health management (PHM), is to identify and anticipate potential issues in the equipment before these become serious. The main challenge to be addressed is to assess the amount of time a piece of equipment will function effectively before it fails, which is known as remaining useful life (RUL). Deep learning (DL) models, such as Deep Convolutional Neural Networks (DCNN) and Long Short-Term Memory (LSTM) networks, have been widely adopted to address the task, with great success. However, it is well known that these kinds of black box models are opaque decision systems, and it may be hard to explain their outputs to stakeholders (experts in the industrial equipment). Due to the large number of parameters that determine the behavior of these complex models, understanding the reasoning behind the predictions is challenging. This paper presents a critical and comparative revision on a number of explainable AI (XAI) methods applied on time series regression models for PM. The aim is to explore XAI methods within time series regression, which have been less studied than those for time series classification. This study addresses three distinct RUL problems using three different datasets, each with its own unique context: gearbox, fast-charging batteries, and turbofan engine. Five XAI methods were reviewed and compared based on a set of nine metrics that quantify desirable properties for any XAI method. One of the metrics introduced in this study is a novel metric. The results show that Grad-CAM is the most robust method, and that the best layer is not the bottom one, as is commonly seen within the context of image processing.

Disparities in Telemedicine Use Among Louisiana Medicaid Beneficiaries During the COVID-19 Pandemic.

The COVID-19 pandemic led to an increased reliance on telemedicine. Whether this exacerbated existing disparities within vulnerable populations is not yet known. Characterize changes in outpatient telemedicine evaluation and management (E&M) services for Louisiana Medicaid beneficiaries by race, ethnicity, and rurality during the COVID-19 pandemic. Interrupted time series regression models estimated pre-COVID-19 trends and changes in E&M service use at the April and July 2020 peaks in COVID-19 infections in Louisiana and in December 2020 after those peaks had subsided. Louisiana Medicaid beneficiaries continuously enrolled between January 2018 and December 2020 who were not also enrolled in Medicare. Monthly outpatient E&M claims per 1000 beneficiaries. Prepandemic differences in service use between non-Hispanic White and non-Hispanic Black beneficiaries narrowed by 34% through December 2020 (95% CI: 17.6%-50.6%), while differences between non-Hispanic White and Hispanic beneficiaries increased by 10.5% (95% CI: 0.1%-20.7%). Non-Hispanic White beneficiaries used telemedicine at higher rates than non-Hispanic Black (difference=24.9 claims per 1000 beneficiaries, 95% CI: 22.3-27.4) and Hispanic beneficiaries (difference=42.3 claims per 1000 beneficiaries, 95% CI: 39.1-45.5) during the first wave of COVID-19 infections in Louisiana. Telemedicine use increased slightly for rural beneficiaries compared with urban beneficiaries (difference=5.3 claims per 1000 beneficiaries, 95% CI: 4.0-6.6). The COVID-19 pandemic narrowed gaps in outpatient E&M service use between non-Hispanic White and non-Hispanic Black Louisiana Medicaid beneficiaries, though gaps in telemedicine use emerged. Hispanic beneficiaries experienced large reductions in service use and relatively small increases in telemedicine use.