Statistical Regression Models Research Articles

The relative effects of access to public greenspace and private gardens on mental health

Although the beneficial effects of urban greenspace on mental health are widely accepted, the comparative effects of public greenspace and private (domestic) gardens on mental health are poorly understood. Here, an assessment is provided of the effect of local public greenspace on a person’s mental health for those with and without a private garden in Britain. Individual level data on private garden ownership and mental health status are obtained from a nationally representative survey (the British Household Panel Survey). A combination of statistical matching and regression models are used to account for individual and area-level confounders and to test for interactions. Individuals with (n = 4,454) and without (n = 338) private gardens are analysed separately and their predicted probability of poor mental health in response to public greenspace presence is compared. Results show that the predicted positive effect of having a private garden varies depending on gender and age. Specifically, having a private garden substantially reduces the maximum predicted probability of poor mental health for men regardless of their access to local public greenspace. Whereas, for women, the presence of local public greenspace results in comparable mental health for those with and without a garden. Women without access to local public greenspace, having a private garden reduces the predicted probability of poor mental health later in life. Given the results, it is recommended that the provision of private gardens is considered within greenspace guidance and policy, which is currently dominated by the provision of, or access to, public greenspaces.

Indoor exposure to polycyclic aromatic hydrocarbons associated with solid fuel use in rural China.

Polycyclic aromatic hydrocarbons (PAHs) are widespread environmental contaminants associated with various health risks including lung cancer. Indoor exposure to PAHs, particularly from the indoor burning of fuels, is significant; however, long-term large-scale assessments of indoor PAHs are hampered by high costs and time-consuming in field sampling and laboratory experiments. A simple fuel-based approach and statistical regression models were developed as a trial to predict indoor BaP, as a typical PAH, in China, and consequently spatiotemporal variations in indoor BaP and indoor exposure contributions were discussed. The results show that the national population-weighted indoor BaP concentration has decreased substantially from 46.1ng/m3 in 1992 to 6.60ng/m3 in 2017, primarily due to the increased use of clean energies for cooking and heating. Indoor BaP exposure contributed to > 70% of the total inhalation exposure in most cities, particularly in regions where solid fuels are widely utilized. With limited experimental observation data in building statistical models, quantitative results of the study are associated with high uncertainties; however, the study undoubtedly supports effective countermeasures on indoor PAHs from solid fuel use and the importance of promoting clean household energy usage to improve household air quality.

Development and evaluation of the MiCheck® Prostate test for clinically significant prostate cancer

BackgroundThere is a clinical need to identify patients with an elevated PSA who would benefit from prostate biopsy due to the presence of clinically significant prostate cancer (CSCaP). We have previously reported the development of the MiCheck® Test for clinically significant prostate cancer. Here, we report MiCheck's further development and incorporation of the Roche Cobas standard clinical chemistry analyzer. ObjectivesTo further develop and adapt the MiCheck® Prostate test so it can be performed using a standard clinical chemistry analyzer and characterize its performance using the MiCheck-01 clinical trial sample set. Design, settings, and participantsAbout 358 patient samples from the MiCheck-01 US clinical trial were used for the development of the MiCheck® Prostate test. These consisted of 46 controls, 137 non-CaP, 62 non-CSCaP, and 113 CSCaP. MethodsSerum analyte concentrations for cellular growth factors were determined using custom-made Luminex-based R&D Systems multi-analyte kits. Analytes that can also be measured using standard chemistry analyzers were examined for their ability to contribute to an algorithm with high sensitivity for the detection of clinically significant prostate cancer. Samples were then re-measured using a Roche Cobas analyzer for development of the final algorithm. Outcome measurements and statistical analysisLogistic regression modeling with Monte Carlo cross-validation was used to identify Human Epidydimal Protein 4 (HE4) as an analyte able to significantly improve the algorithm specificity at 95% sensitivity. A final model was developed using analyte measurements from the Cobas analzyer. ResultsThe MiCheck® logistic regression model was developed and consisted of PSA, %free PSA, DRE, and HE4. The model differentiated clinically significant cancer from no cancer or not-clinically significant cancer with AUC of 0.85, sensitivity of 95%, and specificity of 50%. Applying the MiCheck® test to all evaluable 358 patients from the MiCheck-01 study demonstrated that up to 50% of unnecessary biopsies could be avoided while delaying diagnosis of only 5.3% of Gleason Score (GS) ≥3+4 cancers, 1.8% of GS≥4+3 cancers and no cancers of GS 8 to 10. ConclusionsThe MiCheck® Prostate test identifies clinically significant prostate cancer with high sensitivity and negative predictive value (NPV). It can be performed in a clinical laboratory using a Roche Cobas clinical chemistry analyzer. The MiCheck® Prostate test could assist in reducing unnecessary prostate biopsies with a marginal number of patients experiencing a delayed diagnosis.

Characterizing the spatiotemporal evolution of building material stock in China's Greater Bay Area: A statistical regression method

AbstractMore than half of the materials extracted from natural environments eventually accumulate as building material stock (BMS). From a linear‐to‐circular economy perspective, BMS transforms the building sector from a virgin material consumer and a waste generator to a future depository of secondary resources. Studies characterizing the amount and distribution of BMS adopt different approaches, but high data requirements restrict their applicability. This research proposes an alternative method for regional BMS quantification. The method leverages the permanent population, electricity consumption, and BMS of a sample city to develop a statistical regression model; then uses it to estimate the BMS of a larger, homogenous region. With relatively low data requirements, the new method is especially applicable in underdeveloped areas where data required for BMS quantification methods are usually unavailable or incomplete. We apply the method to characterize the spatiotemporal evolution of BMS in China's Greater Bay Area. From 2000 to 2021, the total BMS in this region increased from 4.4 to 7.7 billion tonnes, with concrete, brick, and steel accounting for 72.32%, 17.57%, and 4.71% of the total BMS, respectively. The most rapid BMS growth occurred in Guangzhou (from 534.75 to 1277.82 Mt) and Shenzhen (517.80 to 1235.48 Mt). A core–edge BMS accumulation pattern emerged in this area while the BMS peak showed a coast‐to‐inland shift. Future studies can explore generalizing this new method to characterize BMS in other developing regions.

Workforce affiliation in primary and secondary prevention implantable cardioverter defibrillator patients: a nationwide Danish study.

There are a paucity of studies investigating workforce affiliation in connection with first-time implantable cardioverter defibrillator (ICD)-implantation. This study explored workforce affiliation and risk markers associated with not returning to work in patients with ICDs. Using the nationwide Danish registers, patients with a first-time ICD-implantation between 2007 and 2017 and of working age (30-65 years) were identified. Descriptive statistic and logistic regression models were used to describe workforce affiliation and to estimate risk markers associated with not returning to work, respectively. All analyses were stratified by indication for implantation (primary and secondary prevention). Of the 4659 ICD-patients of working age, 3300 patients (71%) were members of the workforce (employed, on sick leave or unemployed) (primary: 1428 (43%); secondary:1872 (57%)). At baseline, 842 primary and 1477 secondary prevention ICD-patients were employed. Of those employed at baseline, 81% primary and 75% secondary prevention ICD-patients returned to work within 1 year, whereof more than 80% remained employed the following year. Among patients receiving sick leave benefits at baseline, 25% were employed after 1 year. Risk markers of not returning to work were 'younger age' in primary prevention ICD-patients, while 'female sex', left ventricular ejection fraction 'LVEF≤40', 'lower income', and '≥3 comorbidities' were risk markers in secondary prevention ICD-patients. Lower educational level was a risk marker in both patient groups. High return-to-work proportions following ICD-implantation, with a subsequent high level of employment maintenance were found. Several significant risk markers of not returning to work were identified including 'lower educational level' that posed a risk in both patient groups. Capital Region of Denmark, P-2019-051.

Food Parenting Practices and Child Eating Behaviors in Australian Families: A Cross-Sectional Sibling Design

BackgroundResearch on feeding in early childhood has focused primarily on parent–child dyadic interactions, despite parents enacting these practices within the complex dynamic of the family system. ObjectiveUsing a sibling design, this study aimed to assess how parents may adapt their food parenting practices for siblings in response to differences in their eating behaviors. DesignA cross-sectional online survey was conducted between October and December 2022. Participants/settingData were collected from parents (97.5% women) in Australia with 2 children aged 2 to 5 years (n = 336 parents and n = 672 children). Main outcome measuresSurvey items were completed for each sibling, and included four subscales of the Children's Eating Behaviour Questionnaire and seven subscales of the Feeding Practices and Structure Questionnaire-28. Statistical analyses performedMultiple linear regression models examined associations between within-sibling pair differences in child eating behaviors and food parenting practices, adjusting for differences in child body mass index z score, age, gender, and early feeding method. ResultsWithin-sibling pair differences in eating behaviors were associated with differences in some food parenting practices. For the fussier sibling, parents reported using more control-based practices, including persuasive feeding, reward for eating, and reward for behavior, and less of the structure-based practice, family meal settings (P values < 0.001). Similar directions of associations were found for persuasive feeding, reward for eating, and family meal settings with siblings who were slower eaters or more satiety responsive (P values < 0.007); however, no significant differences in reward for behavior were observed in relation to sibling differences in these eating behaviors. For the more food responsive sibling, parents reported using more control-based practices, including reward for behavior and overt restriction (P values < 0.002). ConclusionsWithin families, parents may adapt certain practices in response to differences in their children's eating behaviors. Interventions promoting responsive feeding should be designed to acknowledge the integral role of siblings in shaping parents' feeding decisions.

Credit risk management and effectiveness of credit activities: Evidence from Vietnam

The research aims to assess the effects of credit risk management on the effectiveness of credit activities at Vietnamese state-owned commercial banks. Based on the literature review and empirical studies, the aspects of credit risk management practices include four factors proposed in the research model, such as: credit terms, client appraisal, credit risk control, and collection policies. The author implements the descriptive statistical method and linear regression model on SPSS with data from a survey of credit officers from state-owned commercial banks in Vietnam. The findings show that credit term, client appraisal, and credit risk control positively influence loan performance at a 95% confidence interval. The paper also highlights the role of internal control as the mechanism for reducing credit risk and improving the effectiveness of credit activities.

An assessment of Sentinel‐1 synthetic aperture radar, geophysical and topographical covariates for estimating topsoil particle‐size fractions

AbstractData derived from synthetic aperture radar (SAR) are widely employed to predict soil properties, particularly soil moisture and soil carbon content. However, few studies address the use of microwave sensors for soil texture retrieval and those that do are typically constrained to bare soil conditions. Here, we test two statistical modelling approaches—linear (with and without interaction terms) and tree‐based models, namely compositional linear regression model (LRM) and random forest (RF)—and both nongeophysical (e.g., surface soil moisture, topographic, etc) and geophysical‐based (electromagnetic, magnetic and radiometric) covariates to estimate soil texture (sand %, silt % and clay %), using microwave remote sensing data (ESA Sentinel‐1). The statistical models evaluated explicitly consider the compositional nature of soil texture and were evaluated with leave‐one‐out cross‐validation (LOOCV). Our findings indicate that both modelling approaches yielded better estimates when fitted without the geophysical covariates. Based on the Nash–Sutcliffe efficiency coefficient (NSE), LRM slightly outperformed RF, with NSE values for sand, silt and clay of 0.94, 0.62 and 0.46, respectively; for RF, the NSE values were 0.93, 0.59 and 0.44. When interaction terms were included, RF was found to outperform LRM. The inclusion of interactions in the LRM resulted in a decrease in NSE value and an increase in the size of the residuals. Findings also indicate that the use of radar‐derived variables (e.g., VV, VH, RVI) alone was not able to predict soil particle size without the aid of other covariates. Our findings highlight the importance of explicitly considering the compositional nature of soil texture information in statistical analysis and regression modelling. As part of the continued assessment of microwave remote sensing data (e.g., ESA Sentinel‐1) for predicting topsoil particle size, we intend to test surface scattering information derived from the dual‐polarimetric decomposition technique and integrate that predictor into the models in order to deal with the effects of vegetation cover on topsoil backscattering.

Development of statistical regression and artificial neural network models for estimating nitrogen, phosphorus, COD, and suspended solid concentrations in eutrophic rivers using UV–Vis spectroscopy

River water quality monitoring is crucial for understanding water dynamics and formulating policies to conserve the water environment. In situ ultraviolet–visible (UV–Vis) spectrometry holds great potential for real-time monitoring of multiple water quality parameters. However, establishing a reliable methodology to link absorption spectra to specific water quality parameters remains challenging, particularly for eutrophic rivers under various flow and water quality conditions. To address this, a framework integrating desktop and in situ UV–Vis spectrometers was developed to establish reliable conversion models. The absorption spectra obtained from a desktop spectrometer were utilized to create models for estimating nitrate-nitrogen (NO3-N), total nitrogen (TN), chemical oxygen demand (COD), total phosphorus (TP), and suspended solids (SS). We validated these models using the absorption spectra obtained from an in situ spectrometer. Partial least squares regression (PLSR) employing selected wavelengths and principal component regression (PCR) employing all wavelengths demonstrated high accuracy in estimating NO3-N and COD, respectively. The artificial neural network (ANN) was proved suitable for predicting TN in stream water with low NH4-N concentration using all wavelengths. Due to the dominance of photo-responsive phosphorus species adsorbed onto suspended solids, PLSR and PCR methods utilizing all wavelengths effectively estimated TP and SS, respectively. The determination coefficients (R2) of all the calibrated models exceeded 0.6, and most of the normalized root mean square errors (NRMSEs) were within 0.4. Our approach shows excellent efficiency and potential in establishing reliable models monitoring nitrogen, phosphorus, COD, and SS simultaneously. This approach eliminates the need for time-consuming and uncertain in situ absorption spectrum measurements during model setup, which may be affected by fluctuating natural and anthropogenic environmental conditions.

Impact of Total Expenses &amp; Labor on the Value of Wheat with a Special Reference to Sustainable Development Goals

Abstract: The framework for sustainable development consists of 17 Sustainable Development Goals (SDGs). The SDGs comprise objectives, benchmarks, and measures for combating poverty, guaranteeing food security, pursuing sustainable agriculture, protecting the environment, encouraging peace, and other issues. One of the most important areas of the Indian economy is agriculture, which provides a living for over 70% of the country's population. In addition, there aren't many job opportunities in other sectors. As a result, it is necessary to study the agriculture sector from the standpoint of sustainable development and make some suggestions for its improvement. This paper develops a statistical regression model to estimate the value of wheat. The study's findings imply that the government should take the required actions to increase farmers' access to funding and raise the standard of the educational programs, training resources, and extension services that are available to them. In order to raise the price of wheat and ultimately their profit, farmers should also invest more money in premium seeds and fertilizers.

Hybrid Teaching Using Problem-Based Learning to Promote Self-Directed Learning Abilities of Students during the COVID-19 Pandemic

This study aims to 1) investigate the self-directed learning (SDL) abilities and factors influencing SDL among undergraduate students, and 2) examine the effects of hybrid teaching using problem-based learning in promoting self-directed learning abilities. The study is divided into two phases. The first phase involves analyzing self-directed learning abilities and factors influencing SDL among students in Thailand Universities, with a sample group of 326 individuals. The research tool in Phase 1 was a Self-Directed Learning evaluation form. The findings reveal that students possess high levels of self-directed learning abilities in all aspects. Furthermore, all factors significantly impacted SDL, with the highest influence in the utilization of information technology. The statistical regression model is represented by "Y" ̂ = 1.542 +0.115X1+0.088X2+0.303X3, indicating that the model can predict the dependent variable with an accuracy of 34%. In the second phase, the effects of the approach in promoting self-directed learning abilities were explored out with an experimental sample group of 17 students, enrolled in the Database Systems for Technology and Computer Innovation course. The main research tools for Phase 2 included (1) a Self-Directed Learning evaluation form, (2) an achievement test, and (3) a questionnaire. The findings revealed the highest level in all aspects, and students who learned through the approach demonstrated higher self-directed learning abilities after the course. Additionally, their post-course learning outcomes were significantly higher than their pre-course outcomes, with statistical significance at the 0.01 level. Students expressed a favorable perception toward the learning approach.

Rice yield prediction using statistical regression models in the selected districts of Maharashtra

Rice yield prediction using statistical regression models in the selected districts of Maharashtra

Spatial differentiation of carbon emissions from energy consumption based on machine learning algorithm: A case study during 2015–2020 in Shaanxi, China

Carbon emissions resulting from energy consumption have become a pressing issue for governments worldwide. Accurate estimation of carbon emissions using satellite remote sensing data has become a crucial research problem. Previous studies relied on statistical regression models that failed to capture the complex nonlinear relationships between carbon emissions and characteristic variables. In this study, we propose a machine learning algorithm for carbon emissions, a Bayesian optimized XGboost regression model, using multi-year energy carbon emission data and nighttime lights (NTL) remote sensing data from Shaanxi Province, China. Our results demonstrate that the XGboost algorithm outperforms linear regression and four other machine learning models, with an R2 of 0.906 and RMSE of 5.687. We observe an annual increase in carbon emissions, with high-emission counties primarily concentrated in northern and central Shaanxi Province, displaying a shift from discrete, sporadic points to contiguous, extended spatial distribution. Spatial autocorrelation clustering reveals predominantly high-high and low-low clustering patterns, with economically developed counties showing high-emission clustering and economically relatively backward counties displaying low-emission clustering. Our findings show that the use of NTL data and the XGboost algorithm can estimate and predict carbon emissions more accurately and provide a complementary reference for satellite remote sensing image data to serve carbon emission monitoring and assessment. This research provides an important theoretical basis for formulating practical carbon emission reduction policies and contributes to the development of techniques for accurate carbon emission estimation using remote sensing data.

Creativity and commerce: a shifting balance for specialty foods and beverages

PurposeProducers of specialty products that apply nonindustrial production techniques to deliver high quality and authenticity to niche markets define themselves in opposition to industrial production. This forces them to strike a delicate balance between, on the one hand, emphasizing the symbolic value related to quality, creativity and distance from traditional industrial production, and on the other hand, ensuring profitability through commercial orientation. This paper aims to explore how new and established firms benefit from a different balance between a strategic orientation toward creativity and the symbolic value of specialty products and commercialization.Design/methodology/approachBased on survey data from 99 specialty coffee roasters, this study uses statistical regression models to analyze how new versus established firms benefit from a strategic orientation toward creativity versus commercialization.FindingsThe authors find that firms benefit from shifting the balance from a strategic orientation on creativity in new firms to a strategic orientation on commercialization in more established firms. This shift increases the likelihood that firms’ passion and creativity lead to commercial success.Originality/valueThe findings demonstrate the importance of strategic orientation and emphasize that business owners need to revisit and adjust their strategic orientation as their businesses develop.

Evaluating the driving forces of spectral inversion methods used for assessing water quality parameters in Poyang Lake, China

AbstractWater quality parameters are key indicators of quality of water and can indicate the algal biomass and eutrophication in lakes. Therefore, this study spectrally inverted and evaluated the water quality of the Poyang Lake in China by analysing the differences between the measured water quality parameters and the observed image spectra of Sentinel‐2 remote sensing. This analysis was done using statistical regression models (SRMs) and various machine learning models (e.g., support vector machine [SVM] and random forest [RF]). The following major conclusions were drawn: (1) nutrients accumulated more in summer (June–August) than in winter (December–February). For example, in local areas of the main river, total nitrogen (TN) concentration reached 3.4 mg/L in summer of 2016, whereas total phosphorus (TP) concentrations were below 0.052 mg/L in winters of 2016 and 2017. (2) The three models, SRM, RF and SVM, achieved good results in the inversions of Secchi depth and permanganate index. However, differences were noted in the inversions of other parameters. For example, the goodness‐of‐fit (r2) between the inversion and measured values of TN concentration from RF was 0.81, while that from SR was 0.73. (3) The spatial distribution patterns of the water quality parameters showed differences. The ammoniacal nitrogen concentration was higher in the central region than that in the western region of the lake, whereas TP concentration was higher at the shoreline of the lake at 0.07 mg/L on 2 April 2017 but was relatively low in the main channel.

Multi-Sensor Fusion Target Tracking Based on Maximum Mixture Correntropy in Non-Gaussian Noise Environments with Doppler Measurements

This paper addresses the multi-sensor fusion target tracking problem based on maximum mixture correntropy in non-Gaussian noise environments exclusively using Doppler measurements. As Doppler measurements are non-linear, a statistical linear regression model is constructed using the unscented transformation. Then, a centralized measurement model is developed, and the mixture correntropy is determined, which contains the high-order statistics of state prediction and the measurement error caused by noise. Then, a robust fusion filter is proposed by maximizing the mixture-correntropy-based cost. To improve numerical stability, the information filter and corresponding square root version are also derived. Furthermore, the performance of the proposed algorithm is analyzed, and the selection of the kernel width is discussed. Experiments are performed using simulated data and automatic driving software. The results show that the estimation performance of the proposed algorithm is better with respect to outliers and mixture Gaussian noise than that of traditional methods.

Hydrogen Peroxide Diffusion through Dental Tissues-In Vitro Study.

Whitening products commonly utilize hydrogen peroxide (HP) as an active principle, which can penetrate dental tissues with potential side effects due to its low molecular weight. This study aimed to evaluate the HP diffusion of two in-office whitening products, namely 6% VivaStyle Paint On Plus (VS) and Opalescence Boost 40% (OP), in different tooth types. Additionally, the influence of the area of exposure, dental tissue thickness and pulp chamber volume was assessed. Each group consisted of eighteen intact anterior (A), premolar (PM) and molar (M) human teeth, and a positive pulpal pressure model was employed. The samples were analyzed using spectrophotometry, and results were expressed as the mean and 95% confidence interval. Statistical tests and linear regression models were appropriately applied at α = 5%. The total HP (µg) retrieved was as follows: VS-A, 1.333 [1.214, 1.452]; OP-A, 1.538 [1.457, 1.620]; VS-PM, 1.208 [1.123, 1.291]; OP-PM, 3.628 [3.401, 3.855]; VS-M, 2.560 [2.297, 2.823]; and OP-M, 4.197 [3.997, 4.396], with statistically significant differences in diffusion kinetics between whitening products for PM and M. Several HP concentrations attained a minimum cytotoxicity value of 2.22 µg/mL. The regression model shows that OP exposed the pulp chamber to 1.421 µg of HP more than that of VS. Different whitening products can cause cytotoxic HP concentrations in the pulp chamber, with a higher risk observed in molars.

Geospatial analytics of driving mechanism of land subsidence in Gulf Coast of Texas, United States

Land subsidence has been an ongoing issue for over a century along the Gulf Coast of Texas in the United States. This study assesses and models the factors contributing to land subsidence covering fifty-six (56) counties along the Gulf of Mexico coastline from Louisiana to the border of Mexico, approximately 300,000 km2. Geospatial statistical techniques and regression models were employed to investigate and predict the fundamental causes of land subsidence by integrating multiple datasets such as Global Navigation Satellite System (GNSS) (147 stations), groundwater extraction (78,420 wells), hydrocarbon production (84,424 wells), precipitation, and population growth. In the last two decades, the overall population rose by 33 % and the compound annual population growth rate increased from 2.08 to 4.10 % in Montgomery, Waller, Fort Bend, and Chambers counties. Emerging hotspot analysis reveals that the groundwater level is persistently declining and the regression model (R2 = 0.92) tested over Harris County predicts that the population growth significantly contributes to land subsidence in this region. The groundwater withdrawal rate is increased from 23 to 96.6 billion gallons in Harris, Montgomery, and Fort Bend counties in the last two decades. A prolonged drought from 2010 to 2015 due to low precipitation affected all fifty-six counties. Oil production increased eightfold and a high extraction rate of 19.5 to 40.1 million bbl/yr of oil in Karnes County was recorded within the last 20 years. The regression model (R2 = 0.73) over this county suggests that oil extraction is a primary contributing factor to the observed subsidence. Although the gas extraction rates for all counties are decreasing over time, some counties in the southern part of the Gulf Coast Aquifer show relatively higher extraction rates. For the first time, this research determines that all fifty-six counties along the Gulf Coast of Texas are undergoing land subsidence and experiencing high population growth, groundwater withdrawal, and hydrocarbon extraction.

Energy-Adjusted Dietary Inflammatory Index and Diabetes Risk in Postmenopausal Hispanic Women

BackgroundType 2 diabetes is a major public health concern in the United States and worldwide. The dietary inflammatory index (DII) and the energy-adjusted DII (E-DII) are tools that assess dietary inflammation. Previous evidence suggests that obesity can modify the association between inflammation and disease. ObjectiveThe aim of this study was to evaluate the association between the DII/E-DII and incident diabetes in self-identified Hispanic women from the Women’s Health Initiative (WHI). The secondary aim was to evaluate whether obesity modifies the association between the DII/E-DII scores and incident diabetes. DesignParticipants were from the WHI Observational Study and the Clinical Trial Components (except women from the treatment arm in the Dietary Modification Trial) conducted among postmenopausal women in the United States. DII/E-DII scores were calculated from a self-administered food frequency questionnaire at baseline that included 122 food items, of which 12 are representative of Hispanic eating patterns. Participants/settingsParticipants included 3,849 postmenopausal women who self-identified as Hispanic that were recruited for the WHI from 1993 to 1998 at 40 US clinical centers. Main outcome measuresThe outcome was incident diabetes. Statistical analysis performedCox regression models were used to assess the association between DII/E-DII and incident diabetes. Models were adjusted for age at baseline, lifestyle-related risk factors, known type 2 diabetes mellitus (T2DM) risk factors, and neighborhood socioeconomic status. Interaction was tested between the DII/E-DII scores and obesity. ResultsThe incidence of diabetes was 13.1% after a median follow-up of 13 years. Higher E-DII scores were associated with a higher risk of incident diabetes (hazard ratio [HR], 1.09; 95% confidence interval [CI], 1.04–1.14). There was no interaction between E-DII scores and obesity (P = 0.73). ConclusionsPro-inflammatory diets, as measured by higher E-DII scores, were associated with a higher risk of incident diabetes. Future research is needed for understanding how the inflammatory potential of diets can be decreased.

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.