Urban Built Environment as a Predictor for Coronary Heart Disease—A Cross-Sectional Study Based on Machine Learning DOI Creative Commons
Dan Jiang, Fei Guo, Ziteng Zhang

и другие.

Buildings, Год журнала: 2024, Номер 14(12), С. 4024 - 4024

Опубликована: Дек. 18, 2024

The relationship between coronary heart disease (CHD) and complex urban built environments remains a subject of considerable uncertainty. development predictive models via machine learning to explore the underlying mechanisms this association, as well formulation intervention policies planning strategies, has emerged pivotal area research. A cross-sectional dataset hospital admissions for CHD over course year from in Dalian City, China, was assembled matched with multi-source environment data residential addresses. This study evaluates five models, including decision tree (DT), random forest (RF), eXtreme gradient boosting (XGBoost), multi-layer perceptron (MLP), support vector (SVM), compares them multiple linear regression models. results show that DT, RF, XGBoost exhibit superior capabilities, all R2 values exceeding 0.70. DT model performed best, an value 0.818, best performance based on metrics such MAE MSE. Additionally, using explainable AI techniques, reveals contribution different factors identifies significant influencing cold regions, ranked age, Digital Elevation Model (DEM), house price (HP), sky view factor (SVF), interaction factors. Stratified analyses by age gender variations groups: those under 60 years old, Road Density is most influential factor; 61–70 group, top 71–80 81 building height leading males, GDP females, factor. explores feasibility predicting risk regions provides comprehensive methodology workflow cardiovascular refined neighborhood-level factors, offering scientific construction sustainable healthy cities.

Язык: Английский

Acute effects of visual and olfactory nature stimuli on task performance DOI Creative Commons

Sarayu Chandramouli,

Suma Katabattuni,

Marco A. Palma

и другие.

Research Square (Research Square), Год журнала: 2024, Номер unknown

Опубликована: Дек. 11, 2024

Abstract Nature exposure enhances human experiences and supports overall wellbeing including physiological (stress), emotional (mood), mental (focus) benefits. However, existing workplace design literature – typically indoors, sensory-restricted, focuses mainly on nature views, despite being a multisensory experience. Rooted in Attention Restoration Theory (ART), Stress Reduction (SRT), "smellscape”, this study implemented between-subjects controlled 2x2 experiment to explore how views scents independently in-combination influence performance, measuring real-effort performance metrics stress, attention, memory recall, reasoning skills, risk aversion, cheating behaviors. Across 256 participants, results show that enhance cognitive boost positive emotions, reduce anxiety. Introducing alongside visual stimuli further amplifies these Physiologically, both reduced blink rates, indicating lower anxiety levels compared nature-absent settings. Our findings highlight the substantial benefits of incorporating passively inexpensively into workplace. By selecting tasks mirror real-world office tasks, advances understanding connection between cognition offices, offering insights creating environments productivity well-being.

Язык: Английский

Процитировано

0
Urban Built Environment as a Predictor for Coronary Heart Disease—A Cross-Sectional Study Based on Machine Learning DOI Creative Commons
Dan Jiang, Fei Guo, Ziteng Zhang

и другие.

Buildings, Год журнала: 2024, Номер 14(12), С. 4024 - 4024

Опубликована: Дек. 18, 2024

The relationship between coronary heart disease (CHD) and complex urban built environments remains a subject of considerable uncertainty. development predictive models via machine learning to explore the underlying mechanisms this association, as well formulation intervention policies planning strategies, has emerged pivotal area research. A cross-sectional dataset hospital admissions for CHD over course year from in Dalian City, China, was assembled matched with multi-source environment data residential addresses. This study evaluates five models, including decision tree (DT), random forest (RF), eXtreme gradient boosting (XGBoost), multi-layer perceptron (MLP), support vector (SVM), compares them multiple linear regression models. results show that DT, RF, XGBoost exhibit superior capabilities, all R2 values exceeding 0.70. DT model performed best, an value 0.818, best performance based on metrics such MAE MSE. Additionally, using explainable AI techniques, reveals contribution different factors identifies significant influencing cold regions, ranked age, Digital Elevation Model (DEM), house price (HP), sky view factor (SVF), interaction factors. Stratified analyses by age gender variations groups: those under 60 years old, Road Density is most influential factor; 61–70 group, top 71–80 81 building height leading males, GDP females, factor. explores feasibility predicting risk regions provides comprehensive methodology workflow cardiovascular refined neighborhood-level factors, offering scientific construction sustainable healthy cities.

Язык: Английский

Процитировано

0