kaceyclougher/Tanzania-water-wells

Classification of operating status for waterwells located in Tanzania. # Predicting Functionality of Tanzanian Water Wells ### Author: Kacey Clougher Within this repository, you will find a comprehensive analysis that classifies Tanzanian water wells into three categories: functional, non-functional, or in need of repair. The aim of this detailed analysis is to ensure accessibility and replicability. ## Repository Structure - Water-Wells-Final.ipynb: Modeling process - EDA-and-Cleaning.ipynb: Exploratory analysis and data cleaning for final data set - README.md: High level README for reviewers of this project - Tanzania Water Wells.pdf: Presentation for use cases ## Project Understanding This project aims to employ machine learning classification models for predicting the operational status of water wells in Tanzania. The classification encompasses three groups: functional, non-functional, and functional but in need of repair. The ultimate goal is to leverage this modeling approach to forecast the functionality of a newly constructed well or the rejuvenation of an existing non-functional well. ## Data Using machine learning classification models to predict the operational status of water wells in Tanzania, we classify each well into three categories: functional, non-functional, and functional but needs repair. You can find the original data set of over 60,000 wells here: [Pump it Up: Data Mining the Water Table] (drivendata.org). Additional features of this dataset included excavation type, age of well, waterpoint type, etc. ## Modeling and Methods The data underwent preprocessing, including feature engineering, filling missing values, and scaling, with the aim of enhancing the accuracy of the model. A logistic regression model was employed to forecast the operational status of water wells in Tanzania. An initial baseline model and a simple first model were developed to establish a foundational understanding of the data. Subsequently, an extensive grid search wa …