Emeriego/PumpHealth

Develop machine learning models to predict the functionality status of water pumps across Tanzania based on a variety of features including location, water quality, management structure, and technical specifications. # WATER PUMP FUNCTIONALITY PREDICTION This project predicts the functionality status of water pumps across Tanzania using machine learning. Based on features such as location, pump type, installer, water quality, and population served, the models classify pumps as functional, functional needs repair, or non-functional. The project includes: Data cleaning and exploratory analysis to understand pump failure patterns Feature engineering and model training for multi-class classification Geospatial visualizations to highlight areas with high failure risk A dashboard for interactive insights and maintenance planning The goal is to support decision-making for water infrastructure management and improve access to clean water in Tanzanian communities. ## List of Columns and their description id – Unique identifier for each waterpoint. amount_tsh – Total static head (volume of water pumped). date_recorded – Date when the waterpoint data was recorded. funder – Organization or individual who funded the pump. gps_height – Elevation of the pump in meters. that is the altitude. installer – Organization or individual who installed the pump. longitude – GPS longitude of the pump location. latitude – GPS latitude of the pump location. wpt_name – Name of the waterpoint. num_private – Number of private water connections. people using it. NB: low number might indicate communal usage like seen in some regions in africa. basin – River basin where the pump is located. subvillage – Subvillage of the pump location. region – Administrative region name. region_code – Numeric code representing the region. district_code – Numeric code for the district. lga – Local government area of the pump. ward – Administrative ward of the pump. population – Number of people served by the pump. public_meeting – Indicates if a public meeting was held for the pump. recorded_by – Name of the organization/person who recorded the data. scheme_management – Organization managing the wate …