Series Of Modeling Techniques Research Articles

1049-P: Finding Equity: Utilizing Artificial Intelligence to Identify Social and Infrastructural Predictors of Diabetes Mortality in Florida

DM outcomes represent one of the largest avoidable cost burdens with opportunity for improvement in the U.S. health care system. Improving health equity in the context of DM will require targeted community improvements, infrastructure investments, and policy interventions that are designed to maximize the impact of resource allocation through the use of available data and computational resources. By using an Artificial Intelligence approach to evaluate over 2000 socio-economic and infrastructural predictors of DM mortality, this study used a specific series of modeling techniques to identify significant predictors without human selection and compare their predictive ability with all possible factors when passed through artificial neural networks. The final regression model using zip code and county level predictors had an R2 of 0.863. Significant predictors included: Population % White, Population % Householders, Population % Spanish spoken at home, Population % Divorced males, Population % With public health insurance coverage, Population % Employed with private health insurance coverage, Manufacturing-Dependent Designation, Low Education Designation, Population % Medicare Part A and B Female Beneficiaries, Number of Short Term General Hospitals with 50-99 Beds. Using a multi-layered perceptron to predict zip codes at risk the C-statistic for all 2000+ predictors was 0.7938 while the 13 selected predictors was 0.8232. This indicates that these factors are highly relevant for DM mortality in Florida. This process was completed without the need of human variable selection and indicates how AI can be used for informative precision public health analyses for targeted population health management efforts. Disclosure A. Cistola: None.

Research on parametric modeling methods for vortex generators on flat plate

Vortex generators (VGs) are increasingly used on wind turbine blades to delay airflow separation. Small-size VGs substantially increase the numerical cost of simulations of large-size wind turbine blades with VGs. A parametric approach is typically used to replace physical VGs with an algorithm in computational fluid dynamics (CFD) research to address this problem. Nevertheless, the accuracy of the parametric approach is determined by the algorithm accuracy for calculating the maximum circulation of the VGs and the methods for modeling the trailing vortex profile. The main objective of this study is to propose a high-accuracy algorithm that describes the coupled responses of neighboring VGs using a series of modeling techniques. Since counter-rotating triangular VGs are typically used in the wind turbine industry, a pair of these devices mounted on a flat plate is investigated using a full-mesh model and a parametric model, considering different incoming velocities, VG incidence angles, and VG spacings. The research results show that the proposed algorithm for max circulation exhibits better agreement with the experimental data than existing algorithms that consider only one VG, and both model types show high consistency in all cases. The results indicate that the proposed parametric approach seems to be more suitable than the single VG algorithms for replacing physical VGs in CFD research of wind turbine blades equipped with VG arrays.

Modelling personal exposure to particulate air pollution: an assessment of time-integrated activity modelling, Monte Carlo simulation & artificial neural network approaches.

An experimental assessment of personal exposure to PM10 in 59 office workers was carried out in Dublin, Ireland. 255 samples of 24-h personal exposure were collected in real time over a 28 month period. A series of modelling techniques were subsequently assessed for their ability to predict 24-h personal exposure to PM10. Artificial neural network modelling, Monte Carlo simulation and time-activity based models were developed and compared. The results of the investigation showed that using the Monte Carlo technique to randomly select concentrations from statistical distributions of exposure concentrations in typical microenvironments encountered by office workers produced the most accurate results, based on 3 statistical measures of model performance. The Monte Carlo simulation technique was also shown to have the greatest potential utility over the other techniques, in terms of predicting personal exposure without the need for further monitoring data. Over the 28 month period only a very weak correlation was found between background air quality and personal exposure measurements, highlighting the need for accurate models of personal exposure in epidemiological studies.