ezgigm/Project3_TanzanianWaterWell_Status_Prediction

Predicting Water Well Functionality in Tanzania. Feature engineering, SMOTE for imbalanced target, tuning with grid search, feature selection for Random Forest, LGBM, XGBoost, k-neighbors... # Pump It Up: Data Mining the Water Table DrivenData has begun a competition 'Pump It Up' to point the Tanzanian clean water problem. In our module 3 project, we worked on this data science competition. **Problem:** Tanzania is the largest country of East-Africa with 59,353,795 population according to worldometers.info. 25 million of this population have lacks access to clean water, 40 million people also have a lack access to improved sanitation. Water is a basic need and for human beings. The Tanzanian Water Ministry agreed with Taarifa and they aimed to solve this problem by improving clean water sources. There are many water wells already established, but some of them are non-functional or needs repair. **Aim:** Our goal in this project is to build a model that predicts the functionality of water points. With this predictive model, authorities can understand which water points are functional, nonfunctional, and functional but it needs to repair. This model can help the Tanzanian government to find likely maintenance needy wells or give useful information for future wells. With this model, we can help the Tanzanian authorities how to use water sources in a productive way. It also helps the investment of the government on wells. **Solution:** With 86% accuracy, our model can predict the functionality of wells. With the good prediction of functionality, the solutions can be; - prioritizing functioning wells which need repair and yield clean water - targeting repairs to clusters of wells especially those with high populations - payments of some kind will provide an incentive to keep wells functional - allocate funds and resources to effective organizations with a track record **Data:** The original data was obtained from the DrivenData 'Pump it Up: Data Mining the Water Table' competition. Basically, there are 4 different data sets; submission format, training set, test set and train labels set which contains status of wells. With given training set and la …