nyambura-tech/Water-Pump-Functionality-Prediction

Tanzania faces critical challenges with water access as thousands of waterpoints are non-functional, leaving communities without reliable water. This project builds a machine learning model to predict waterpoint operational status, enabling the government and NGOs to prioritize maintenance and allocate resources efficiently. # Water Pump Functionality Prediction A supervised machine learning project to classify the operational condition of waterpoints across Tanzania. ## Table of Contents - Problem Statement - Dataset Overview - Type of ML Problem - Project Folder Structure - ML Workflow - Model Deployment with Flask - Notes & Hints - Deliverables Checklist ## Problem Statement Access to clean and functional water is a critical challenge across many parts of Tanzania. Thousands of waterpoints, wells, boreholes, communal standpipes and more have been installed across the country, but a significant number are either non-functional or in need of repair, leaving communities without reliable water access. ### Objective Build a machine learning model that predicts the operational condition of a waterpoint given a set of features describing its location, management, water source and technical setup. Accurate predictions can help the Tanzanian government and NGOs prioritise maintenance efforts and allocate resources more efficiently. ### Target Variable The model must classify each waterpoint into one of three categories: | Label | Description | |-------|-------------| | functional | The waterpoint is operational and needs no repairs | | functional needs repair | The waterpoint is operational but requires repairs | | non functional | The waterpoint is not operational | ## Dataset Overview **Important Note on Dataset Files:** The dataset is split across two separate files: - `train.csv` - Contains all the input features for each waterpoint - `labels.csv` - Contains the target column (`status_group`) for each waterpoint Both files share a common `id` column. You must merge/join them on `id` as your very first step before any exploration or modelling. ### Feature Columns | Column | Description | Example Value | |--------|-------------|---------------| | amount_tsh | Total static head — amount of water available to the waterpoint | 300.0 | | date_recorded | Date the row was entered i …