farzaneh-gerami/HydroLogic

Machine learning project to classify water pump functionality in Tanzania based on location, water quality, management, and technical data — supporting better infrastructure planning and water access. # HydroLogic **HydroLogic** is a machine learning project focused on predicting the functionality of water pumps in Tanzania. Using features like location, water quality, and technical specifications, the goal is to classify pumps as: - **Functional** - **Functional Needs Repair** - **Non-Functional** This project is inspired by the DrivenData: Pump it Up competition and aims to improve maintenance planning and water access across communities. --- ## Project Structure HydroLogic/ │ ├── data/ # Raw and processed datasets ├── notebooks/ # Jupyter notebooks for EDA and modeling ├── src/ # Scripts for preprocessing and modeling ├── dashboard/ # Streamlit dashboard (coming later) ├── requirements.txt # Python dependencies ├── .gitignore # Files/folders to exclude from version control └── README.md # You're here! ## Project Goals Perform data cleaning and feature engineering Build classification models to predict pump status Visualize insights with geospatial plots Deploy an interactive dashboard using Streamlit ## Weekly Milestones Week 1: Project setup, initial data exploration Week 2-3: Data cleaning, feature engineering, EDA Week 4-6: Modeling and optimization Week 7-8: Dashboard development and deployment Week 9: Final presentation and documentation ## Team Roles Data Preprocessing Lead: [Name] Modeling Specialist: [Name] Dashboard Developer: [Farzaneh Gerami] ## Acknowledgments Data: Taarifa & Tanzania Ministry of Water Competition: DrivenData - Pump it Up