Time Series Regression Models Research Articles

Seasonal Variation in Neurologic Hospitalizations in the United States.

Certain neurologic diseases have been noted to vary by season, and this is important for understanding disease mechanisms and risk factors, but seasonality has not been systematically examined across the spectrum of neurologic disease, and methodologic guidance is also lacking. Using nationally representative data from the National Inpatient Sample, a stratified 20% sample of all non-federal acute care hospitalizations in the United States, we calculated the monthly rate of hospitalization for 14 neurologic diseases from 2016 to 2018. For each disease, we assessed seasonality of hospitalization using chi-squared, Edward, and Walter-Elwood tests and seasonal time series regression models. Statistical tests were adjusted for multiple hypothesis testing using Bonferroni correction. Meningitis, encephalitis, ischemic stroke, intracerebral hemorrhage, Guillain-Barre syndrome, and multiple sclerosis had statistically significant seasonality according to multiple methods of testing. Subarachnoid hemorrhage, status epilepticus, myasthenia gravis, and epilepsy had significant seasonality according to Edwards and Walter-Elwood tests but not chi-square tests. Seasonal time series regression illustrated seasonal variation in all 14 diseases of interest, but statistical testing for seasonality within these models using the Kruskal-Wallis test only achieved statistical significance for meningitis. Seasonal variation is present across the spectrum of acute neurologic disease, including some conditions for which seasonality has not previously been described, and can be examined using multiple different methods. ANN NEUROL 2023;93:743-751.

Main drivers of diabetes pharmaceuticals expenditures: evidence from OECD countries and Iran.

This study aimed to identify the impact of prominent drivers on drug expenditure for diabetes. Following the examination of previous studies, this study identified possible factors contributing to diabetes pharmaceutical expenditures. The explanatory variables for the study were the median population age, access to innovative drugs, GDP per capita, prevalence, price, and consumption of diabetes drugs. Then, to estimate the per capita expenditure among diabetic patients, this study developed the panel data model and two time-series regression models for OECD countries and Iran, respectively. In the panel data regression model, R2 was 0.43. The influence of the age, prevalence, consumption volume and GDP per capita coefficients were + 1.79, + 0.704, + 3.86057, + 0.00054, respectively. Also, the probability level of all variables was less than 0.05. In Iran's comparative time-series regression model, R2 was 0.9, and the only significant influence coefficient was the age (β=+0.91). In the another model for Iran, R2 was 0.99, the influence coefficient of age was + 0.249, the prevalence was + 0.131, innovation was + 0.029, and the price was + 0.00054; all the probability levels were less than 0.05. Pharmaceutical per capita expenditure is affected by several factors. These factors are not the same in various counties. Passing a judgment on drug utilization only based on pharmaceutical per capita expenditure cannot be perfect. Also, judging whether the per capita drug expenditure in one country is desirable without attention to the affecting factors and only relying on the value of utilized medicines is defective.

Universal electronic consultation program (e-consultation) of a cardiology department: long-term results in mortality

Abstract Background Many health systems have initiated electronic consultation programs, although little is known about their impact on accessibility, safety and satisfaction (1). The implementation of integrated electronic medical record projects between healthcare levels allows all healthcare professionals access to all the clinical information of patients, which is a key factor in optimizing the management of healthcare resources, facilitating the communication between professionals and avoiding medical acts that do not add value, always from the premise of safety and quality for patients (2). In our model, a General Practitioner sends an e-referral to the Cardiology Department (CD). A Cardiologist can either answer without needing the face-to-face consultation, or schedule the patient for a single act consultation. In our health area, from 2009 to 2012, a “single act” cardiology consultation model was implemented, according to which the appointment includes, on the same day, all the complementary tests that might be necessary for the diagnosis. From 2013 until present, an electronic consultation or e-consultation model has been implemented; consisting of a telematic evaluation of the e-consultation by a cardiologist, who can either solve the GP's requests directly or book an appointment for the patient in person, in cases where it is considered necessary (3). Purpose The aim of this study is to assess the clinical impact of the implementation of a model of outpatient care that includes an initial electronic consultation (e-consultation), comparing it with a face-to-face single act model. Methods We selected patients with at least one visit to the Cardiology Department between 2010 and 2019. Using an interrupted time series regression model, we analysed the impact of incorporating the e-consultation into the healthcare model (started in 2013), evaluating: waiting time for care and mortality. Results We analysed 47,377 patients: 61.9% attended by incorporating the e-consultation and 38.1% in the face-to-face consultation model. The delay time for care in the e-consultation model (median [IQR]: 7 days [5–13]) was shorter than in the face-to-face model (median [IQR]: 33 days [14–81]), p<0.001. The regression model for interrupted time series showed that the incorporation of e-consultation contributed a very important decrease in the delay of attention, remaining around 9 days, although with slight oscillations, Figure 1. The patients evaluated via e-consultation had less mortality (2.5% vs 3.9%, p<0.001), Figure 2. Conclusions An outpatient care program that includes an e-consultation has shown to reduce waiting times significantly and is a safe model with a lower rate of mortality over the first year after the consultation. Funding Acknowledgement Type of funding sources: None.

An Empirical Analysis of the Impact of Global Financial Crisis on Economic Activity in North Macedonia

Global financial crisis caused major economic disruptions and slow down economic activity in North Macedonian economy. The purpose of this paper is to examine the impact of Global financial crisis on economic activity in North Macedonia through analyzing fluctuations of major macroeconomic indicators of the economy. In this regard, it analyzes the trends of macroeconomic indicators that reflect the development and macroeconomic balances of those economies, such as GDP, unemployment, foreign direct investment, trade openness, and the level of public debt, in order to identify factors that have contributed in their fluctuations. The research methodology consists of regression model of time series, OLS and VECM models. The results suggest that Global financial crisis had a negative impact on the economic growth, because the coefficient has a negative sign and is statistically significant. Keywords: global financial crisis, economic effects, OLS, VECM DOI: 10.7176/EJBM/14-23-01 Publication date: December 31 st 2022

A Forecast of Surface Ozone Using Analytical Models

In this study, several analytical models were tested to forecast the surface ozone concentration using Artificial Neural Network (ANN), Multiple Linear Regression (MLR) and Time Series Regression (TSR). Four study areas were selected in this study, namely Seberang Jaya in Penang, Shah Alam in Selangor, Larkin in Johor and Kota Bharu in Kelantan. The main objective of this study is to determine the appropriate analytical models MLR and ANN for surface ozone forecasting in some zones of peninsular Malaysia, to forecast surface ozone concentration with TSR model in several zones of peninsular Malaysia and to compare the performance of each model by the performance index. The performance index that will be shown in this study for the model comparison are root mean square error (RMSE), mean square error (MSE) and determination of coefficient (R2). The ANN model showed better performance compared to the MLR and TSR models in the model comparison in each station. The station in Larkin, Johor provides high accuracy in forecasting surface ozone concentrations for each model with minimum MSE, 0.000009 ppm and RMSE, 0.0042 ppm compared to other stations. The value of R2 is 0.33 which is highest compared to station in Seberang Jaya and Kota Bharu.

A Predictive Analysis of Retail Sales Forecasting using Machine Learning Techniques

In a retail industry, sales forecasting is an important part related to supply chain management and operations between the retailer and manufacturers. The abundant growth of the digital data has minimized the traditional system and approaches to do a specific task. Sales forecasting is the most challenging task for the inventory management, marketing, customer service and Business financial planning for the retail industry. In this paper we performed predictive analysis of retail sales of Citadel POS dataset, using different machine learning techniques. We implemented different regression (Linear regression, Random Forest Regression, Gradient Boosting Regression) and time series models (ARIMA LSTM), models for sale forecasting, and provided detailed predictive analysis and evaluation. The dataset used in this research work is obtained from Citadel POS (Point Of Sale) from 2013 to 2018 that is a cloud base application and facilitates retail store to carryout transactions, manage inventories, customers, vendors, view reports, manage sales, and tender data locally. The results show that Xgboost outperformed time series and other regression models and achieved best performance with MAE of 0.516 and RMSE of 0.63.

The impact of carbon finance on energy consumption structure: evidence from China.

As a financial innovation model to promote sustainable economic development, carbon finance promotes green and inclusive economic development by reshaping the financial system. As a big energy consumer, China has been committed to the green transformation of energy structure, and carbon finance has been regarded as an important tool to support carbon emission reduction. This paper divides carbon finance from three levels, financial macro, environmental governance, and technological development, and selects 17 indexes to measure the development level of carbon finance in China. Time series regression model and the entropy weight method (EWM) are used to analyze the impact of carbon finance on China's energy consumption structure. The empirical results show that carbon finance plays a positive role in promoting China's energy consumption structure, and developing carbon finance is conducive to the use of clean energy and the optimization of energy consumption structure in China; carbon finance can further optimize China's energy consumption structure by investing capital, resources, and technology in industries related to promoting clean energy consumption. Therefore, China should continue to develop carbon finance and promote the establishment of a unified national carbon trading market; at the same time, China should use the CCER trading mechanism to promote the development of renewable energy and enhance the endogenous driving force of energy transformation by promoting technological progress.

P099Statewide changes in service trends at utah title x clinics around the domestic gag rule, 2017–2021

<h3>Objectives</h3> Utah is one of six states that lost all Title X funding with the passage of the Domestic Gag Rule in July 2019. We explored trends in service provision at Utah Title X clinics from 2017 to 2021 to evaluate the impact of the loss of this funding. <h3>Methods</h3> Using service delivery data from seven Title X clinics collected between January 2017 and August 2021, we identified most frequently used diagnostic and procedural codes and compared trends in service provision and service utilization using interrupted time series analyses (ITSA), risk ratios, and regression models. <h3>Results</h3> We saw decreases in average patient age and higher proportions of patients with public/private insurance compared to self-payers each year. Type of care provided changed significantly over time: odds decreased that a visit was related to contraception (OR, 0.86; 95% CI, 0.85–0.86) or cervical cancer screening (OR, 0.81; 95% CI, 0.80–0.82), and increased that a visit included sexually transmitted infection (STI) screening (OR, 1.10; 95% CI, 1.10–1.11). Despite these trends, our ITSA found the impact of the Gag Rule to be nonsignificant atop the already decreasing trend in monthly contraceptive visits (p=0.58). <h3>Conclusions</h3> Despite concerns that loss of Title X funding would decrease contraception provision, our analyses identified a shift in care provision away from contraception and preventive health, towards more acute services like STI testing and treatment. Trend analysis showed this shift beginning as early as 2017, with no significant level change after the Gag Rule was enacted. This suggests that other factors influenced service changes.

Effect size quantification for interrupted time series analysis: implementation in R and analysis for Covid-19 research

BackgroundInterrupted time series (ITS) analysis is a time series regression model that aims to evaluate the effect of an intervention on an outcome of interest. ITS analysis is a quasi-experimental study design instrumental in situations where natural experiments occur, gaining popularity, particularly due to the Covid-19 pandemic. However, challenges, including the lack of a control group, have impeded the quantification of the effect size in ITS. The current paper proposes a method and develops a user-friendly R package to quantify the effect size of an ITS regression model for continuous and count outcomes, with or without seasonal adjustment.ResultsThe effect size presented in this work, together with its corresponding 95% confidence interval (CI) and P-value, is based on the ITS model-based fitted values and the predicted counterfactual (the exposed period had the intervention not occurred) values. A user-friendly R package to fit an ITS and estimate the effect size was developed and accompanies this paper. To illustrate, we implemented a nation population-based ITS study from January 2001 to May 2021 covering the all-cause mortality of Israel (n = 9,350 thousand) to quantify the effect size of Covid-19 exposure on mortality rates. In the period unexposed to the Covid-19 pandemic, the mortality rate decreased over time and was expected to continue decreasing had Covid-19 not occurred. In contrast, the period exposed to the Covid-19 pandemic was associated with an increased all-cause mortality rate (relative risk = 1.11, 95% CI = 1.04, 1.18, P < 0.001).ConclusionFor the first time, the effect size in ITS: was quantified, can be estimated by end-users with an R package we developed, and was demonstrated with data showing an increase in mortality following the Covid-19 pandemic. ITS effect size reporting can assist public health policy makers in assessing the magnitude of the entire intervention effect using a single, readily understood measure.

Community-in-the-loop: Creating Artificial Process Intelligence for Co-production of City Service

Communities have first-hand knowledge about community issues. This study aims to improve the efficiency of social-technical problem-solving by proposing the concept of "artificial process intelligence," based on the theories of socio-technical decision-making. The technical challenges addressed were channeling the communication between the internal-facing and external-facing 311 categorizations. Accordingly, deep learning models were trained on data from Kansas City's 311 system: (1) Bidirectional Encoder Representations from Transformers (BERT) based classification models that can predict the internal-facing 311 service categories and the city departments that handle the issue; (2) the Balanced Latent Dirichlet Allocation (LDA) and BERT clustering (BLBC) model that inductively summarizes residents' complaints and maps the main themes to the internal-facing 311 service categories; (3) a regression time series model that can predict response and completion time. Our case study demonstrated that these models could provide the information needed for reciprocal communication, city service planning, and community envisioning. Future studies should explore interface design like a chatbot and conduct more research on the acceptance and diffusion of AI-assisted 311 systems.

Excess suicide attributable to the COVID-19 pandemic and social disparities in South Korea

The impact of COVID-19 pandemic on suicide remains unclear and might differ according to individuals’ socioeconomic characteristics. We aimed to investigate excess suicide attributable to COVID-19 in South Korea, stratified by the outbreak period and individual characteristics. We obtained daily time-series suicide mortality data for January 2017–December 2020 from the Korea National Statistics Office and performed a two-stage interrupted time-series analysis. We estimated excess suicide in 16 regions of Korea using a quasi-Poisson time-series regression model and pooled the region-specific estimates using a mixed-effects multivariate meta-analysis model in the first and second stages, respectively. From February 18 to December 31, 2020, suicide decreased by 9.5% [95% empirical confidence interval (eCI): 3.8%, 15.6%] compared to the number expected from the pre-pandemic period. The decrease in excess suicide risk from the initial pandemic was pronounced during the pandemic’s first and third waves. Further, we found that the decrease in suicide was more evident in individuals who were male [11.7% (95% eCI: 5.5%, 18.0%)], middle-aged [13.7% (95% eCI: 7.8%, 19.6%)], highly educated [12.6% (95% eCI: 6.4%, 19.4%)], and married [13.6% (95% eCI: 8.0%, 20.3%)] than in the general population, based on the point estimates. Our results provide timely evidence to establish public health policies for suicide prevention and suggest the prioritization of resource allocation for mental health of individuals based on individual characteristics.

Examining the Effectiveness of the 2012 Canadian Graphic Warning Label Policy Change by Sex, Income, and Education.

We examined the differential impact of the 2012 Canadian GWL policy changes on key indicators of warning label impact and quit intentions using national cohorts of Canadian and U.S. adults who smoke. We used data from all waves of the International Tobacco Control surveys (2002-2020) in Canada and the United States. Our key measures were quit intentions and an index of warning label effectiveness (salience, cognitive and behavioral reactions). We estimated overall policy impact by comparing Canada (treatment group) with the United States (control group) using controlled interrupted time series (CITS) regression models, with interactions to examine whether policy impact varied by sex, education, and income. The CITS model showed a statistically significant increase in the warning label effectiveness in Canada post-policy, compared to the United States (β = 0.84, 95% CI 0.35,1.33). Similarly, the odds of quit intentions were relatively higher among adults who smoked in Canada compared to the United States (OR = 1.89, 95% CI 1.51,2.36) post-policy. The three-way interaction model showed that these associations were greater among adults from low socioeconomic status (SES) groups than in high SES groups. The 2012 change in the Canadian GWL policy was associated with stronger cognitive and behavioral responses to GWLs and higher odds of quit intentions among adults who smoked in Canada when compared to the United States, specifically among individuals from low SES groups, suggesting a positive equity impact. Our findings affirm the need for countries to implement or enhance GWLs, in line with the WHO Framework Convention on Tobacco Control (FCTC). The evidence on the potential health equity benefit of GWL policies is mixed. To further understand the influence of GWL policies on tobacco use disparities, more systematic research using pre/post-policy designs with control groups is needed. Using a CITS model, we aimed to strengthen the available evidence on the causal influence of this tobacco control approach. Our findings show that the 2012 GWL policy change had a greater impact on adults who smoked from low SES groups than it did on adults who smoked from high SES groups, indicating a potentially positive equity impact and confirming the need for countries to implement or maximize the size of GWLs, as recommended by the WHO FCTC.

Universal electronic consultation program (e-consultation) of a cardiology department: long-term results in mortality

Abstract Background Many health systems have initiated electronic consultation programs, although little is known about their impact on accessibility, safety and satisfaction (1). The implementation of integrated electronic medical record projects between healthcare levels allows all healthcare professionals access to all the clinical information of patients, which is a key factor in optimizing the management of healthcare resources, facilitating the communication between professionals and avoiding medical acts that do not add value, always from the premise of safety and quality for patients (2). In our model, a General Practitioner sends an e-referral to the Cardiology Department (CD). A Cardiologist can either answer without needing the face-to-face consultation, or schedule the patient for a single act consultation. In our health area, from 2009 to 2012, a “single act” cardiology consultation model was implemented, according to which the appointment includes, on the same day, all the complementary tests that might be necessary for the diagnosis. From 2013 until present, an electronic consultation or e-consultation model has been implemented; consisting of a telematic evaluation of the e-consultation by a cardiologist, who can either solve the GP's requests directly or book an appointment for the patient in person, in cases where it is considered necessary (3). Purpose The aim of this study is to assess the clinical impact of the implementation of a model of outpatient care that includes an initial electronic consultation (e-consultation), comparing it with a face-to-face single act model. Methods We selected patients with at least one visit to the Cardiology Department between 2010 and 2019. Using an interrupted time series regression model, we analysed the impact of incorporating the e-consultation into the healthcare model (started in 2013), evaluating: waiting time for care and mortality. Results We analysed 47,377 patients: 61.9% attended by incorporating the e-consultation and 38.1% in the face-to-face consultation model. The delay time for care in the e-consultation model (median [IQR]: 7 days [5–13]) was shorter than in the face-to-face model (median [IQR]: 33 days [14–81]), p&amp;lt;0.001. The regression model for interrupted time series showed that the incorporation of e-consultation contributed a very important decrease in the delay of attention, remaining around 9 days, although with slight oscillations, Figure 1. The patients evaluated via e-consultation had less mortality (2.5% vs 3.9%, p&amp;lt;0.001), Figure 2. Conclusions An outpatient care program that includes an e-consultation has shown to reduce waiting times significantly and is a safe model with a lower rate of mortality over the first year after the consultation. Funding Acknowledgement Type of funding sources: None.

Development of summer skiing days in Austrian glacier ski areas in the first two decades of the twenty-first century.

Summer skiing on glaciers is a leisure activity highly dependent on natural factors like weather and glacier conditions. While the literature agrees that climate change is shaping the fate of summer skiing, longer time series covering both the supply and demand side of glacier/summer skiing, and research that combines natural and non-natural factors affecting the summer ski market are missing. To close this research gap, we conducted a detailed supply-side analysis of nine Austrian glacier ski areas focusing on the number of operating days (in the summer half-year, as well as for meteorological and astronomical summer ski definitions), show how these numbers evolve over time, test how they correlate with meteorological and glaciological data using time-series regression models on a yearly and monthly basis, and highlight how managerial decisions come into play when trying to explain the summer ski supply’s decline. Between 2002 and 2019, summer ski operating days in Austrian glacier ski areas declined by 48.3% in the summer half-year, 65.2% in the meteorological, and 62.3% in the astronomical summer parallel to rising mean temperatures and shrinking glaciers. This decrease is strongest in June to September and weakest in May and October but in two glacier ski areas, the operating days in the summer season remain constant or are even on the rise. This is in line with model results indicating that meteorological and glaciological data only explain parts of the variance of the decline trends. Operators’ agency, strategies, and decisions play an important role underlining that global warming is not monocausally determining summer ski operation.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00484-022-02371-6.

Fama and French (2015) five-factor Model with Mediating Role of Consumer Confidence: Evidence from Non-financial Sectors of Pakistan

Using a sizable sample of data from the Pakistan Stock Exchange, the study explores the relationship between five risk indicators and portfolio stock returns as well as the mediating function of consumer confidence. The analysis is based on a five-factor model developed by Fama and French (2015) that incorporates time-series ordinary least square regression and structural equation modelling approaches. Market estimates show a statistically significant relationship with portfolio excess returns in terms of size, value, profitability, and investment risk-premia. Nevertheless, Consumer Confidence, a mediating and independent variable in Pakistani non-financial firms, is insignificant in capturing the variation of portfolio returns.

Testing the Validity of the Unconditional and Conditional CAPM on the Egyptian Stock Market Using Panel Data Analysis

The purpose of this study is to test the validity of the capital asset pricing model (CAPM) in the Egyptian stock market, in its unconditional and conditional forms. Both simple and multiple panel data analyses are used to test the CAPM, using data from 118 individual stocks and 10 stock portfolios over the 2011-2019 period, based on 1062 firm/year observations and 90 portfolio/year observations for each stock market index. Four stock market indices are used to estimate beta using five-year monthly data, based on 4248 time series regression models for the period 2006- 2018. The results support the conditional CAPM, which considers market condition, and contradict the unconditional one. These results are robust to the use of other data frequencies (daily and weekly) to estimate beta, and to the exclusion of financial sector firms. The main practical implication of this study is to guide Firms’ stakeholders such as investors, portfolio managers, firm managers, and researchers to employ the conditional CAPM in predicting stock returns based on panel data analysis. Further, several avenues for future research are proposed.

GDP Economic Forecasting Model Based on Improved RBF Neural Network

Among the existing GDP forecasting methods, time series forecasting and regression model forecasting are the two most commonly used forecasting methods. However, traditional macroeconomic forecasting models are unable to accurately achieve optimal forecasts of highly complex nonlinear dynamic macroeconomic systems due to the influence of multiple confounding factors. In order to solve the above problems, a GDP economic forecasting model based on an improved RBF neural network is proposed. First, the main traditional GDP forecasting methods are analyzed. Then, RBF neural networks are used to solve the problem that traditional forecasting technology methods cannot handle multi-factor complex nonlinearities well. Second, to further improve the convergence speed and accuracy of the RBF neural network learning algorithm, the Shuffled Frog Leaping Algorithm with global search capability and high practicality is fused into the RBF network training. Finally, the improved RBF neural network is used to build a GDP economic forecasting model. The performance of the Shuffled Frog Leaping Algorithm and the improved RBF neural network was tested using the approximation of Hermit polynomials and the Iris classification problem as simulation examples. The experimental results show that the improved RBF neural network-based GDP economic forecasting model achieves more accurate forecasting accuracy than other forecasting methods.

Simple statistical models can be sufficient for testing hypotheses with population time-series data.

Time‐series data offer wide‐ranging opportunities to test hypotheses about the physical and biological factors that influence species abundances. Although sophisticated models have been developed and applied to analyze abundance time series, they require information about species detectability that is often unavailable. We propose that in many cases, simpler models are adequate for testing hypotheses. We consider three relatively simple regression models for time series, using simulated and empirical (fish and mammal) datasets. Model A is a conventional generalized linear model of abundance, model B adds a temporal autoregressive term, and model C uses an estimate of population growth rate as a response variable, with the option of including a term for density dependence. All models can be fit using Bayesian and non‐Bayesian methods. Simulation results demonstrated that model C tended to have greater support for long‐lived, lower‐fecundity organisms (K life‐history strategists), while model A, the simplest, tended to be supported for shorter‐lived, high‐fecundity organisms (r life‐history strategists). Analysis of real‐world fish and mammal datasets found that models A, B, and C each enjoyed support for at least some species, but sometimes yielded different insights. In particular, model C indicated effects of predictor variables that were not evident in analyses with models A and B. Bayesian and frequentist models yielded similar parameter estimates and performance. We conclude that relatively simple models are useful for testing hypotheses about the factors that influence abundance in time‐series data, and can be appropriate choices for datasets that lack the information needed to fit more complicated models. When feasible, we advise fitting datasets with multiple models because they can provide complementary information.

Analysis System of MICE Tourism Economic Development Strategy Based on Machine Learning Algorithm

The environmental effect of the meetings, incentives, conventions, and exhibitions (MICE) industry is as extensive as its economic impact. Visitors attending events use a wide range of service providers, including airline car rental firms, restaurants, hotels, theaters, and tour operators. Traditionally used tourism demand forecasting approaches rely heavily on univariate time series and multivariate regression models. Although these function-based prediction systems have demonstrated some effectiveness in forecasting tourism, they are unable to accurately capture the link between tourist demand and supply as a feed-forward neural network does (FFNN). Research has shown that an FFNN can outperform regression and time-series algorithms when it comes to forecasting tourism data. This research, for the first time, expands the use of neural networks in tourist demand creation by combining a hybrid FFNN and chimp optimization learning algorithm (i.e., FFNN-ChOA) into a nonlinear tourism demand dataset. In terms of predicting accuracy, FFNN-ChOA surpasses traditional backpropagation neural networks, regression models, and time-series models.

Associations of inclement weather and poor air quality with non-motorized trail volumes

Inclement weather and poor air quality present unfavorable conditions for active transportation. This study investigated associations of area-wide air pollution and inclement weather with non-motorized travel at two trail locations in northern Utah (one recreational, one utilitarian). Rainfall, snowfall, high/low temperatures, and high wind speed were measures of inclement weather, and overall air quality index (AQI) measured air pollution. This study fit time-series (ARIMA) regression models for the utilitarian and recreational sites (separately), using multiple years of daily non-motorized counts. Non-motorized travel was reduced significantly during inclement weather, and the effects of rainfall and cold temperatures were smaller for the utilitarian site than for the recreational site. Counts at the recreational site were higher on holidays, weekends, and days with greater snow depth. Although non-motorized counts were lower on days with worse air quality at the utilitarian location, they were higher at the recreational location. Results are interpreted in terms of the discretionary nature of recreational travel, particulars of the location and elevation of the recreational site, and a desire to escape pollution through outdoor recreation in a natural area.