anewt225/tanzania_water

Flatiron Mod 3 Tanzania Water Dataset # Tanzania Water Flatiron Mod 3 Project ## Overview ### Repository Navigation - Data Folder Includes - Data from Taarifa on datadriven.org - Training Set Labels - Training Set Values - Test Set Values - Modified Datasets - Combined Training Set - Altitude Data - Cleaned Data - Preliminary Data - Notebooks: Summary Notebook - Exploratory Data Analysis - Data Production - Determine Confidence Intervals - Train Regression - Make Prediction - Visualizations - Presentation Presentation ### ReadMe Navigation 1. Business Context 2. Current Landscape 1. The problem 2. Data Understanding 3. Data Limitations 3. Predictive Analysis 1. Performance Metric 2. Model Selection 3. Feature Importance 4. Conclusion & Takeaways 1. Improvement Areas 2. Growth Opportunities 3. Takeaways 5. Conclusions 6. Further Steps 7. Project Info *** ## Business Context Water in Africa: Technical and Equipment Researchers, LTD. (WATER) is a (hypothetical) consultancy which seeks to amplify the positive effects of progress in water point parts, supply, and maintenance. Progress may be made by the public and/or private sectors by analyzing, and anticipating the status of water point repair more quickly and efficiently. ### Predicting Water Point Maintenance Status from Data The instrumental goal in this project is to construct a model capable of predicting the status of a waterpoint (specifically, in Tanzania), between the three states of "functional", "functional needs repair", and "non functioning".The underlying motivation is to create a cheaply implemented, readily available tool for businesses or other organizations looking to provide maintenance and replacement services for said waterpoints. This should in principle lower the overhead costs for such providers (less chance of over-ordering parts, over-hiring staff, etc.) and so should, due to lower volatility and increased profitability, stimulate the creation of new providers or expansion of new ones. A major factor in waterpo …