PaulaOrdonez10/taarifa-water-pump-prediction

Machine Learning project predicting water pump functionality in Tanzania using classification models and feature importance analysis. # Taarifa Water Pump Prediction ## Overview This project focuses on predicting the operational status of water pumps in Tanzania using supervised Machine Learning techniques. The objective is to classify water pumps into three categories: - Functional - Non Functional - Functional Needs Repair The project was developed as part of the Master's in Data Science, Big Data & Business Analytics at Universidad Complutense de Madrid. --- ## Technologies Used - Python - Pandas - NumPy - Scikit-Learn - Matplotlib - Seaborn --- ## Machine Learning Workflow - Data Cleaning - Exploratory Data Analysis (EDA) - Missing Value Treatment - Feature Engineering - Categorical Encoding - Model Validation - Model Comparison - Prediction Generation - Competition Submission --- ## Models Evaluated Several supervised Machine Learning algorithms were tested and compared: - Random Forest Classifier - Balanced Random Forest - Extra Trees Classifier - HistGradientBoosting Classifier The final model was selected based on validation performance and competition score. --- ## Final Results The best performing model was an optimized Random Forest classifier, achieving a competition score of **0.8191**. The analysis also identified the most influential variables affecting water pump functionality. ### Top Features Key drivers included: - Longitude - Latitude - Water Quantity - GPS Height - Construction Year - Population - Waterpoint Type - Funder Information --- ## Skills Demonstrated - Classification Modelling - Feature Engineering - Missing Value Treatment - Feature Importance Analysis - Model Evaluation - Hyperparameter Optimization - Predictive Analytics - Machine Learning Pipelines --- ## Repository Structure ```text notebooks/ └── ucm_taarifa_final.ipynb images/ └── feature_importance.png data/ └── Dataset files are not included if they exceed GitHub upload limits. ``` --- ## Author Paula Ordóñez Montoya Master's in Data Science, Big Data & Business Analytics …