pavsob/Maintenance-of-waterpoints---Neural-Nets

Build and trained neural network that helps to improve the maintenance operations at waterpoints in Tanzania making the water more accessible and clean . # Maintenance of waterpoints - Neural Network A neural network was built and trained to help improve the maintenance operations at waterpoints in Tanzania and make the water more accessible and clean. Access to safe water is one of the most critical necessities and a very basic human right. However, according to a report by the World Bank in 2018, “more than 23 million citizens in Tanzania retrieve drinking water from unimproved sources, and 41 million people use unimproved sanitation facilities”. In many rural areas of the country, the water pump systems are poorly maintained, resulting in low-quality water or no access to water. This had negatively affected many people’s health and quality of life. This project investigates building a neural network that detects the water pumps that need to be repaired. It was separated into four tasks. The first starts with a very simple setting where we only consider numeric input features and with balanced data (Task 1). In Task 2, all provided input features were considered (but still with balanced data). Task 3 and Task 4 investigated and solved issues with imbalanced data. You can read more in the report.