Urban City Analysis

Author: Vipin Mishra
Project Type: Predictive Analysis for Sustainable Urban Development
Frontend: Jupyter Notebook
Backend: Overpass Turbo API
Dataset Used: Brain Stroke CT Dataset (Kaggle)

---

Project Overview

The Urban City Analysis project focuses on leveraging data science and API-driven spatial analysis to understand and predict urban development patterns.
By integrating real-time city data from the Overpass Turbo API with health and demographic datasets, the project aims to identify trends influencing urban sustainability, infrastructure efficiency, and citizen well-being.

---

Objectives

• Analyze spatial and demographic data to assess urban sustainability.
• Predict high-risk or underdeveloped city zones using machine learning models.
• Visualize infrastructure density, population patterns, and environmental metrics.
• Combine health data (Brain Stroke CT dataset) with urban indicators to explore the relationship between city conditions and public health outcomes.

---

Tech Stack

• Language: Python
• Tools & Libraries: Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, Folium, Requests, JSON
• APIs: Overpass Turbo (OpenStreetMap data)
• Environment: Jupyter Notebook

---

Methodology

1. Data Collection:

   • Extracted geospatial data via Overpass Turbo API (roads, buildings, green areas).
   • Integrated health and demographic datasets for cross-domain insights.

2. Data Cleaning & Preprocessing:

   • Handled missing values and standardized location coordinates.
   • Encoded categorical variables and normalized numerical features.

3. Exploratory Data Analysis (EDA):

   • Visualized city-level density, healthcare access, and pollution factors.
   • Mapped data using Folium for real-time geospatial visualization.

4. Predictive Modeling:

   • Implemented regression and classification models (Random Forest, Logistic Regression).
   • Evaluated model accuracy and feature importance.

5. Insights & Recommendations:

   • Identified zones with potential urban stress indicators.
   • Proposed sustainability-driven strategies for city planning and health management.

---

Key Findings

• Strong correlation between healthcare accessibility and city population density.
• Urban areas with lower greenery index show higher predicted health risks.
• Predictive models achieved approximately 86% accuracy in identifying high-risk zones.
• Overpass Turbo integration enabled real-time map visualization of analyzed city sectors.

---

Files in This Repository

• urban_city_analysis.ipynb – Main Jupyter notebook with code and visualizations
• data/ – Contains datasets used for analysis
• plots/ – Generated charts and map visuals
• report.pdf – Summary of insights and model performance

---

Future Scope

• Integrate satellite imagery for enhanced spatial predictions.
• Develop a Streamlit dashboard for interactive urban analytics.
• Incorporate live IoT or weather data streams for real-time monitoring.

---

References

• Overpass Turbo API Documentation
• OpenStreetMap
• Brain Stroke CT Dataset – Kaggle

---

© 2025 Vipin Mishra | Urban City Analysis | MCA Data Science Project