scholarly journals Adjusting for non-response in the Finnish Drinking Habits Survey

2019 ◽  
Vol 47 (4) ◽  
pp. 469-473 ◽  
Author(s):  
Hanna Tolonen ◽  
Miika Honkala ◽  
Jaakko Reinikainen ◽  
Tommi Härkänen ◽  
Pia Mäkelä
Keyword(s):  
Logistic Regression ◽  
Random Forest ◽  
Alcohol Use ◽  
Population Level ◽  
Added Value ◽  
Drinking Habits ◽  
Wide Range ◽  
Random Forest Models ◽  
Survey Estimates ◽  
Design Weights

Aim: We aim to compare four different weighting methods to adjust for non-response in a survey on drinking habits and to examine whether the problem of under-coverage of survey estimates of alcohol use could be remedied by these methods in comparison to sales statistics. Method: The data from a general population survey of Finns aged 15–79 years in 2016 ( n=2285, response rate 60%) were used. Outcome measures were the annual volume of drinking and prevalence of hazardous drinking. A wide range of sociodemographic and regional variables from registers were available to model the non-response. Response propensities were modelled using logistic regression and random forest models to derive two sets of refined weights in addition to design weights and basic post-stratification weights. Results: Estimated annual consumption changed from 2.43 litres of 100% alcohol using design weights to 2.36–2.44 when using the other three weights and the estimated prevalence of hazardous drinkers changed from 11.4% to 11.4–11.8%, correspondingly. The use of weights derived by the random forest method generally provided smaller estimates than use of the logistic regression-based weights. Conclusions: The use of complex non-response weights derived from the logistic regression model or random forest are not likely to provide much added value over more simple weights in surveys on alcohol use. Surveys may not catch heavy drinkers and therefore are prone for under-reporting of alcohol use at the population level. Also, factors other than sociodemographic characteristics are likely to influence participation decisions.

scholarly journals Identification of Variable Importance for Predictions of Mortality From COVID-19 Using AI Models for Ontario, Canada

2021 ◽  
Vol 9 ◽  
Author(s):  
Brett Snider ◽  
Edward A. McBean ◽  
John Yawney ◽  
S. Andrew Gadsden ◽  
Bhumi Patel
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Logistic Regression ◽  
Random Forest ◽  
Census Data ◽  
Age Groups ◽  
Area Under The Curve ◽  
Medical Systems ◽  
Wide Range ◽  
Artificial Neural

The Severe Acute Respiratory Syndrome Coronavirus 2 pandemic has challenged medical systems to the brink of collapse around the globe. In this paper, logistic regression and three other artificial intelligence models (XGBoost, Artificial Neural Network and Random Forest) are described and used to predict mortality risk of individual patients. The database is based on census data for the designated area and co-morbidities obtained using data from the Ontario Health Data Platform. The dataset consisted of more than 280,000 COVID-19 cases in Ontario for a wide-range of age groups; 0–9, 10–19, 20–29, 30–39, 40–49, 50–59, 60–69, 70–79, 80–89, and 90+. Findings resulting from using logistic regression, XGBoost, Artificial Neural Network and Random Forest, all demonstrate excellent discrimination (area under the curve for all models exceeded 0.948 with the best performance being 0.956 for an XGBoost model). Based on SHapley Additive exPlanations values, the importance of 24 variables are identified, and the findings indicated the highest importance variables are, in order of importance, age, date of test, sex, and presence/absence of chronic dementia. The findings from this study allow the identification of out-patients who are likely to deteriorate into severe cases, allowing medical professionals to make decisions on timely treatments. Furthermore, the methodology and results may be extended to other public health regions.

scholarly journals Binge Drinking among Residential Program Youth in Israel: Toward an Understanding of Predictors for Policy and Prevention

2018 ◽  
Vol 1 (1) ◽  
pp. 110-127
Author(s):  
Richard Isralowitz ◽  
Alexander Reznik ◽  
Masood Zangeneh
Keyword(s):  
High School ◽  
Alcohol Use ◽  
Binge Drinking ◽  
Prescription Drugs ◽  
Systems Approach ◽  
Residential Program ◽  
Alcohol Availability ◽  
Drinking Habits ◽  
Wide Range ◽  
Drinking Age

Alcohol use is attributed to about 25% of the total deaths among youth and young adults. Harmful alcohol use among youth has been overshadowed by the preoccupation with widespread use of other substances including cannabis and prescription drugs. A crosssectional cohort of 1,327 residential program and high school youth were compared regarding binge drinking habits and risk factors. Data was collected from 2004 to 2016. Residential program youth binge drinking predictors were substance abuse within the last month, alcohol availability, causing harm to others (e.g., fighting, stealing and possessing a weapon), unstructured day activity, and being a passenger in a car where the driver had been drinking. Binge drinking predictors among high school were smoking within the last month and alcohol availability. Effective risk behavior prevention involves a wide range of factors including the need to control alcohol access among those under the legal drinking age. An eco-systems approach involving youth and people they are in contact with is a viable prevention approach. However, conflicting personal and economic factors regarding alcohol use, among others, are a daunting barrier to overcome.

scholarly journals Development of global temperature and pH calibrations based on bacterial 3-hydroxy fatty acids in soils

2020 ◽  
Author(s):  
Pierre Véquaud ◽  
Sylvie Derenne ◽  
Alexandre Thibault ◽  
Christelle Anquetil ◽  
Giuliano Bonanomi ◽  
...  
Keyword(s):  
Fatty Acids ◽  
Random Forest ◽  
Linear Regression ◽  
Membrane Lipids ◽  
Global Scale ◽  
Hydroxy Fatty Acids ◽  
Linear Regression Models ◽  
Wide Range ◽  
Forest Models ◽  
Random Forest Models

Abstract. 3-hydroxy fatty acids (3-OH FAs) with 10 to 18 C atoms are membrane lipids mainly produced by Gram-negative bacteria. They have been recently proposed as temperature and pH proxies in terrestrial settings. Nevertheless, the existing correlations between pH/temperature and indices derived from 3-OH FA distribution (RIAN, RAN15 and RAN17) are based on a small soil dataset (ca. 70 samples) and only applicable regionally. The aim of this study was to investigate the applicability of 3-OH FAs as mean annual air temperature (MAAT) and pH proxies at the global level. This was achieved using an extended soil dataset of 168 topsoils distributed worldwide, covering a wide range of temperatures (5 °C to 30 °C) and pH (3 to 8). The response of 3-OH FAs to temperature and pH was compared to that of established branched GDGT-based proxies (MBT'5Me/CBT). Strong linear relationships between 3-OH FA-derived indices (RAN15, RAN17 and RIAN) and MAAT/pH could only be obtained locally, for some of the individual transects. This suggests that these indices cannot be used as paleoproxies at the global scale using simple linear regression models, in contrast with the MBT'5Me and CBT. However, strong global correlations between 3-OH FA relative abundances and MAAT/pH were shown by using other algorithms (multiple linear regression, k-NN and random forest models). The applicability of the k-NN and random forest models for paleotemperature reconstruction was tested and compared with the MAAT record from a Chinese speleothem. The calibration based on the random forest model appeared to be the most robust. It showed similar trends with previously available records and highlighted known climatic events poorly visible when using local 3-OH FA calibrations. Altogether, these results demonstrate the potential of 3-OH FAs as paleoproxies in terrestrial settings.

scholarly journals Landslide Susceptibility Mapping of Karakorum Highway Combined with the Application of SBAS-InSAR Technology

Sensors ◽  
2019 ◽  
Vol 19 (12) ◽  
pp. 2685 ◽  
Author(s):  
Fumeng Zhao ◽  
Xingmin Meng ◽  
Yi Zhang ◽  
Guan Chen ◽  
Xiaojun Su ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Random Forest ◽  
Landslide Susceptibility ◽  
Susceptibility Mapping ◽  
Normal Operation ◽  
Landslide Susceptibility Mapping ◽  
Geological Conditions ◽  
Random Forest Models ◽  
Sbas Insar ◽  
Very High

Geological conditions along the Karakorum Highway (KKH) promote the occurrence of frequent natural disasters, which pose a serious threat to its normal operation. Landslide susceptibility mapping (LSM) provides a basis for analyzing and evaluating the degree of landslide susceptibility of an area. However, there has been limited analysis of actual landslide activity processes in real-time. The SBAS-InSAR (Small Baseline Subsets-Interferometric Synthetic Aperture Radar) method can fully consider the current landslide susceptibility situation and, thus, it can be used to optimize the results of LSM. In this study, we compared the results of LSM using logistic regression and Random Forest models along the KKH. Both approaches produced a classification in terms of very low, low, moderate, high, and very high landslide susceptibility. The evaluation results of the two models revealed a high susceptibility of land sliding in the Gaizi Valley and the Tashkurgan Valley. The Receiver Operating Characteristic (ROC) curve and historical landslide verification points were used to compare the evaluation accuracy of the two models. The Area under Curve (AUC) value of the Random Forest model was 0.981, and 98.79% of the historical landslide points in the verification points fell within the range of high and very high landslide susceptibility degrees. The Random Forest evaluation results were found to be superior to those of the logistic regression and they were combined with the SBAS-InSAR results to conduct a new LSM. The results showed an increase in the landslide susceptibility degree for 2808 cells. We conclude that this optimized landslide susceptibility mapping can provide valuable decision support for disaster prevention and it also provides theoretical guidance for the maintenance and normal operation of KKH.

scholarly journals Predicting lying, sitting, walking and running using Apple Watch and Fitbit data

2021 ◽  
Vol 7 (1) ◽  
pp. e001004
Author(s):  
Daniel Fuller ◽  
Javad Rahimipour Anaraki ◽  
Bongai Simango ◽  
Machel Rayner ◽  
Faramarz Dorani ◽  
...  
Keyword(s):  
Random Forest ◽  
Population Level ◽  
Wearable Devices ◽  
Support Vector ◽  
Convenience Sample ◽  
Rotation Forest ◽  
Metabolic Equivalents ◽  
Forest Models ◽  
Random Forest Models ◽  
Movement Type

ObjectivesThis study’s objective was to examine whether commercial wearable devices could accurately predict lying, sitting and varying intensities of walking and running.MethodsWe recruited a convenience sample of 49 participants (23 men and 26 women) to wear three devices, an Apple Watch Series 2, a Fitbit Charge HR2 and iPhone 6S. Participants completed a 65 min protocol consisting of 40 min of total treadmill time and 25 min of sitting or lying time. The study’s outcome variables were six movement types: lying, sitting, walking self-paced and walking/running at 3 metabolic equivalents of task (METs), 5 METs and 7 METs. All analyses were conducted at the minute level with heart rate, steps, distance and calories from Apple Watch and Fitbit. These included three different machine learning models: support vector machines, Random Forest and Rotation forest.ResultsOur dataset included 3656 and 2608 min of Apple Watch and Fitbit data, respectively. Rotation Forest models had the highest classification accuracies for Apple Watch at 82.6%, and Random Forest models had the highest accuracy for Fitbit at 90.8%. Classification accuracies for Apple Watch data ranged from 72.6% for sitting to 89.0% for 7 METs. For Fitbit, accuracies varied between 86.2% for sitting to 92.6% for 7 METs.ConclusionThis preliminary study demonstrated that data from commercial wearable devices could predict movement types with reasonable accuracy. More research is needed, but these methods are a proof of concept for movement type classification at the population level using commercial wearable device data.

scholarly journals Predicting the need for intubation in the first 24 h after critical care admission using machine learning approaches

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Benjamin Ming Kit Siu ◽  
Gloria Hyunjung Kwak ◽  
Lowell Ling ◽  
Pan Hui
Keyword(s):  
Machine Learning ◽  
Logistic Regression ◽  
Critical Care ◽  
Random Forest ◽  
Critically Ill ◽  
Critically Ill Patients ◽  
Learning Approaches ◽  
Safety Margins ◽  
Critical Care Admission ◽  
Random Forest Models

AbstractEarly and accurate prediction of the need for intubation may provide more time for preparation and increase safety margins by avoiding high risk late intubation. This study evaluates whether machine learning can predict the need for intubation within 24 h using commonly available bedside and laboratory parameters taken at critical care admission. We extracted data from 2 large critical care databases (MIMIC-III and eICU-CRD). Missing variables were imputed using autoencoder. Machine learning classifiers using logistic regression and random forest were trained using 60% of the data and tested using the remaining 40% of the data. We compared the performance of logistic regression and random forest models to predict intubation in critically ill patients. After excluding patients with limitations of therapy and missing data, we included 17,616 critically ill patients in this retrospective cohort. Within 24 h of admission, 2,292 patients required intubation, whilst 15,324 patients were not intubated. Blood gas parameters (PaO2, PaCO2, HCO3−), Glasgow Coma Score, respiratory variables (respiratory rate, SpO2), temperature, age, and oxygen therapy were used to predict intubation. Random forest had AUC 0.86 (95% CI 0.85–0.87) and logistic regression had AUC 0.77 (95% CI 0.76–0.78) for intubation prediction performance. Random forest model had sensitivity of 0.88 (95% CI 0.86–0.90) and specificity of 0.66 (95% CI 0.63–0.69), with good calibration throughout the range of intubation risks. The results showed that machine learning could predict the need for intubation in critically ill patients using commonly collected bedside clinical parameters and laboratory results. It may be used in real-time to help clinicians predict the need for intubation within 24 h of intensive care unit admission.

DSM-5 alcohol use disorder features among treatment-seeking older adults

2018 ◽  
Vol 64 (4) ◽  
pp. 185-196 ◽  
Author(s):  
Silke Behrendt ◽  
Barbara Braun ◽  
Randi Bilberg ◽  
Gerhard Bühringer ◽  
Michael Bogenschutz ◽  
...  
Keyword(s):  
Older Adults ◽  
Alcohol Use ◽  
Alcohol Use Disorder ◽  
The Elderly ◽  
Treatment Seeking ◽  
Controlled Clinical Trial ◽  
Dsm 5 ◽  
Wide Range ◽  
And Gender ◽  
International Multicenter

Abstract. Background: The number of older adults with alcohol use disorder (AUD) is expected to rise. Adapted treatments for this group are lacking and information on AUD features in treatment seeking older adults is scarce. The international multicenter randomized-controlled clinical trial “ELDERLY-Study” with few exclusion criteria was conducted to investigate two outpatient AUD-treatments for adults aged 60+ with DSM-5 AUD. Aims: To add to 1) basic methodological information on the ELDERLY-Study by providing information on AUD features in ELDERLY-participants taking into account country and gender, and 2) knowledge on AUD features in older adults seeking outpatient treatment. Methods: baseline data from the German and Danish ELDERLY-sites (n=544) were used. AUD diagnoses were obtained with the Mini International Neuropsychiatric Interview, alcohol use information with Form 90. Results: Lost control, desired control, mental/physical problem, and craving were the most prevalent (> 70 %) AUD-symptoms. 54.9 % reported severe DSM-5 AUD (moderate: 28.2 %, mild: 16.9 %). Mean daily alcohol use was 6.3 drinks at 12 grams ethanol each. 93.9 % reported binging. More intense alcohol use was associated with greater AUD-severity and male gender. Country effects showed for alcohol use and AUD-severity. Conclusion: European ELDERLY-participants presented typical dependence symptoms, a wide range of severity, and intense alcohol use. This may underline the clinical significance of AUD in treatment-seeking seniors.

Sign in / Sign up

Export Citation Format

Share Document