Logo Lanfrica

Andrew-Carl/Pump-it-Up-Data-Mining-the-Tanzania-Water-Table

Domaine:

environment and energysocioeconomic

Type de record:

dataset
Créateur:
And
Hôte:
Predicting functionality of groundwater pumps throughout Tanzania. # Pump it Up, Data Mining the Tanzania Water Table ** Photo from(flickr.com) # Background: Currently, the people of Tanzania have poor access to clean drinking water throughout the entire country. Approximately 47% of all Tanzanian citizens do not have access to clean drinking water. Over $1.4 billion dollars in foreign aid has been giving to Tanzania in an attempt to help fix the freshwater crisis. However, the Tanzanian government is failing to fix this problem. A good proportion of the water pumps are completely non-functioning or barely functional and also in need of repair. Many people are left to drink dirty, pathogen filled water, or walk miles on end to the closest functional ground water pump. # Task/Goal - Use machine learning models to predict the functionality of all ground water pumps found throughout the country of Tanzania. - If models are accurate, this could help save the Tanzanian government a lot of time and money. - Accurate models can help to cut the cost on needing workers drive out to every water pump to inspect them. - The government can use this study to know exactly which pumps are working, need repair and which ones aren’t working at all. # How the data was acquired: This project was for a competition on datadriven.org. The website describes itself as: "DrivenData works on projects at the intersection of data science and social impact, in areas like international development, health, education, research and conservation, and public services." Companies, non-profits and governments of high paid prizes in order to have data scientist compete in order to figure out real world problems.The data for this project was acquired from the Tanzania Ministry of Water and the Taarifa Platform. The Taarifa Platform is an open source API designed to use citizen feedback on local problems. The competition and the data can be found at the link below: drivendata.org