IrunguShawn/Tanzania-Wellwatch-ML-Prediction

This is a hackathon competition ### Project Description ## TANZANIA WELL WATCH ML PREDICTION ### AUTHOR - SHAWN J. IRUNGU WellWatch ML is a machine learning model for classifying water well functionality in Tanzania, originally developed as part of Taarifa Hackathon hosted at Driven Data where it placed 796 out of 18,373 competitors. This project served as a showcase of my ability to rapidly prototype impactful data science solutions under tight constraints, leveraging techniques like ML methods and use of specific key tools, e.g., Python, Scikit-learn, preprossessing to achieve performance metrics in predicting well status. This project demonstrates my end-to-end problem-solving skills, from hackathon-speed iteration to production-ready refinement. ### Overview This project seeks to build a Machine Learning classifier algorithm that can predict the condition of a water well (functional, functional-but-needs-repair, and non-functional), using data such as the kind of pump, when it was installed, the installer, the region, and so on. ### Business Understanding Tanzania is a country in East Africa known for its national parks and wild animals. The World Bank estimates its population at 65 million as of 2022 and its land size is about 947,303 km2 (365,756 sq mi). The country comprises many lakes, national parks, and in fact, Africa's highest point, Mount Kilimanjaro (5,895 m or 19,341 ft). Northeast Tanzania is mountainous, while the central area is part of a large plateau covered in grasslands. However, like many other sub-Saharan African countries, Tanzania is a developing country struggling to provide adequate clean water for its bulging population that is growing at 3% per annum. Extreme poverty rate was 44% in 2022 as the population continued to grow at 3% per year. This means an increasing population is not affording the basic quality of life. #### GROUND WATER SITUATION IN TANZANIA According to Dr. Japhet J. Kashaigili the hydrogeology of Tanzania has not bee …