Urban City Analysis ML

Urban City Analysis ML is a comprehensive, menu-driven data science solution for analyzing urban city data using machine learning. The project enables you to predict crime occurrence, accident severity, and passenger count from real-world urban datasets with detailed explanations and user-friendly interface.

---

🚀 Main Features

📊 Comprehensive Data Analysis Pipeline

• Smart Data Preprocessing: Automatically handles missing values, encodes categorical features, and scales numerical data
• Detailed Statistics Display: Shows dataset overview with unique values, ranges, and data quality metrics
• Exploratory Data Analysis: Interactive visualizations including box plots, correlation heatmaps, and scatter plots

🎯 Three ML Analysis Options

1. 🚨 Crime Prediction Analysis

• Purpose: Predicts likelihood of crime occurrence based on urban environmental factors
• Algorithm: Ensemble stacking model combining Random Forest and Gradient Boosting
• Features Used:
  • Passenger count and trip duration
  • Geographic location (district, latitude, longitude)
  • Weather conditions and vehicle count
  • Time-based features (hour, day of week, weekend indicator)
• Output: Accuracy metrics, feature importance, confusion matrix, and classification report

2. 🚗 Accident Severity Prediction

• Purpose: Classifies accident severity levels (low/medium/high)
• Algorithms: Multiple classification models (Random Forest, Decision Tree, Logistic Regression)
• Features Used: Weather conditions, vehicle count, location data, traffic factors
• Output: Comparative model performance, best model selection, and detailed metrics

3. 🚌 Passenger Count Prediction

• Purpose: Estimates passenger count using advanced regression analysis
• Algorithm: Random Forest Regression with hyperparameter tuning via GridSearchCV
• Features Used: Weather conditions, time factors, stop locations, accident severity, crime category
• Advanced Features: Feature engineering with interaction terms and cross-validation
• Output: R² scores, MSE, RMSE, MAPE, and feature importance rankings

💫 User Experience Features

• Menu-Driven Interface: Easy navigation with colored, descriptive menu options
• Progress Indicators: Real-time feedback with emojis and colored status messages
• Detailed Explanations: Each analysis includes methodology descriptions and result interpretations
• Error Handling: Graceful handling of missing data and user input validation
• Optional Visualizations: Choose whether to display data exploration charts

---

🛠️ How to Run

1. Clone the Repository

git clone https://github.com/dharm1123/Urban-City-Analysis-ML.git
cd Urban-City-Analysis-ML

2. Install Required Libraries

pip install -r requirements.txt

Main Dependencies:

• pandas (data manipulation)
• numpy (numerical computations)
• scikit-learn (machine learning algorithms)
• matplotlib & seaborn (data visualization)
• termcolor (colored terminal output)

3. Prepare the Dataset

• Ensure final_crime_dataset.csv is in the project root directory
• For passenger count analysis, passenger_count_dataset_modified.csv is also recommended
• The system will automatically handle missing files and adapt accordingly

4. Run the Main Program

python PROJECTEND.py

5. Follow the Interactive Menu

1. Data Loading: The program automatically loads and preprocesses your data
2. Visualization Choice: Decide whether to see exploratory data analysis charts
3. Analysis Selection: Choose from:
   • 1 - Crime Prediction Analysis
   • 2 - Accident Severity Prediction
   • 3 - Passenger Count Prediction
   • 0 - Exit
4. Results Review: Examine detailed performance metrics and interpretations

---

📁 File Structure

Urban-City-Analysis-ML/
├── PROJECTEND.py                          # 🎯 Main program (menu-driven interface)
├── final_crime_dataset.csv               # 📊 Primary dataset (user provided)
├── passenger_count_dataset_modified.csv  # 🚌 Passenger count dataset (optional)
├── requirements.txt                       # 📦 Python dependencies
├── README.md                             # 📚 This documentation
├── About project.docx                    # 📄 Project overview document
├── model.docx                            # 📈 Model documentation
└── [Additional model scripts]            # 🔧 Individual analysis scripts

---

🔬 Technical Details

Data Preprocessing Pipeline

1. Missing Value Handling: Smart imputation using random sampling from existing values
2. Feature Engineering: Time-based features, interaction terms, categorical encoding
3. Data Scaling: StandardScaler for numerical features, one-hot encoding for categorical
4. Data Quality Checks: Outlier detection and removal, data type validation

Model Performance Metrics

• Classification: Accuracy, Precision, Recall, F1-Score, Confusion Matrix
• Regression: R² Score, MSE, RMSE, MAPE, Cross-validation scores
• Feature Analysis: Feature importance rankings and interpretations

Advanced Features

• Ensemble Methods: Stacking classifier for improved prediction accuracy
• Hyperparameter Tuning: GridSearchCV for optimal model parameters
• Cross-Validation: Robust model evaluation with multiple data splits
• Balanced Datasets: Automatic handling of imbalanced data through sampling

---

👥 Authors & Contributors

Primary Author: DHARM DUDHAGARA
Contributors:

• DHARM DUDHAGARA
• Deep Isalaniya

---

📞 Contact & Support

For questions, suggestions, feature requests, or technical support:

• GitHub Issues: Report bugs or request features
• Repository: View source code and documentation

---

🌟 Key Improvements in This Version

• ✅ Menu-driven interface for better user experience
• ✅ Comprehensive explanations for each analysis type
• ✅ Error handling and validation for robust operation
• ✅ Detailed progress indicators with colored output
• ✅ Advanced feature engineering and model optimization
• ✅ Professional documentation with clear methodology descriptions
• ✅ Modular code structure for easier maintenance and extension

---

Built with ❤️ for urban data science and smart city analysis