meljoy1099/waterwell-status-prediction

Predicting the functionality of waterpoints in Tanzania to help solve their water crisis. Modeling techniques include Logistic Regression, K Nearest Neighbors, Decision Trees, Random Forest, and XG Boost. # Tanzanian Water Wells Status Prediction **Author**: Melody Bass ## Overview Tanzania is a developing country that struggles to get clean water to its population of 59 million people. According to WHO, 1 in 6 people in Tanzania lack access to safe drinking water and 29 million don't have access to improved sanitation. The focus of this project is to build a classification model to predict the functionality of waterpoints in Tanzania given data provided by Taarifa and the Tanzanian Ministry of Water. The model was built from a dataset containing information about the source of water and status of the waterpoint (functional, functional but needs repairs, and non functional) using an iterative approach and can be found here. The dataset contains 60,000 waterpoints in Tanzania and the following features will be used in our final model: * amount_tsh - Total static head (amount water available to waterpoint) * gps_height - Altitude of the well * installer - Organization that installed the well * longitude - GPS coordinate * latitude - GPS coordinate * basin - Geographic water basin * region - Geographic location * population - Population around the well * recorded_by - Group entering this row of data * construction_year - Year the waterpoint was constructed * extraction_type_class - The kind of extraction the waterpoint uses * management - How the waterpoint is managed * payment_type - What the water costs * water_quality - The quality of the water * quantity - The quantity of water * source_type - The source of the water * waterpoint_type - The kind of waterpoint The first sections focus on investigating, cleaning, wrangling, and reducing dimensionality for modeling. The next section contains 6 different classification models and evaluation of each, ultimately leading to us to select our best model for predicting waterpoint status based on the precision of the functional wells in the model. Finally, I will make recommendations to the Tanzanian Government and prov …