Urban Analytics Coursework - Exploring Relationships between Physical Urban Environment and Mental Health
This notebook serves as coursework for the Python Tool for Urban Analytics course at the University of Glasgow.
Throughout the course, I acquired a solid foundation and comprehensive knowledge of various data science techniques, including data visualization, data mining, statistical modeling, machine learning, and deep learning using the Python programming language.
For this coursework, I leveraged my background in urban planning and concentrated on my personal interest in contributing to the public good by utilizing spatial data. The primary goal was to explore the relationships between the physical urban environment and mental health problems at the neighborhood level.
Drawing on urban planning theory, I incorporated various urban environment variables into the analysis. These included neighborhood socioeconomic status (SES), libraries, bus stop density, museums, retail density, distance to general practitioners (GPs), and leisure centers, among others.
To address spatial autocorrelation observed in linear regression models, I introduced the Geographical Weighted Regression (GWR) model. This approach effectively captured spatial heterogeneity in the data, resulting in an improved adjusted R square. The findings indicated that the association between neighborhood characteristics and mental health varies across the study region. This suggests that certain neighborhood characteristics may exert a more significant influence on mental health in specific areas compared to others.
The variations observed in these associations may be attributed to additional personal or environmental factors, necessitating further analysis in future research.
Feel free to explore the notebook for detailed insights and findings.
Note: This project is part of the coursework and focuses on a specific aspect of urban analytics and spatial data analysis.