Preface: -------- This is a part of my contribution to the Kiva, to help them continue and expand their initiative to alleviate global poverty. Context: -------- In order for Kiva to best set its investment priorities, help inform lenders, and understand their target communities, knowing the level of poverty for each borrower is crucial. However, attaining individual-level information is time-consuming, labor-intensive and expensive. Therefore, I propose a method that combines machine learning and satellite imagery to predict poverty. This approach is easier, less expensive and scalable since satellite data is often inexpensive and open source. Before developing this model we have to first collect the data that would be relevant for prediction regionalized poverty, which is why I have created this dataset. Content: -------- This data set is a collection of the following: 1. Environmental Data: Vegetation indices, soil characteristics, evaporation 2. Climate Data: Temperature, precipitation, elevation 3. Socioeconomic and Demographic Data: Population Density, Access to major cities, Nightlight, Land usage 4. Conflict Data: Conflicts, death tolls, civilian casualties 5. Natural Disaster Data (coming shortly) I have provided the data in the following formats: - Individual Level Data: For each region in the loans (kiva_loans) and MPI study set, I have extracted all of the information listed above. The data is provided in the csv files; MPIData_augmented.csv, kivaData_augmented.csv - Stacked satellite images for each country with a loan. For each country, a satellite image is provided as a .grd file in the Satellite Imagery folder Usage: ------ If you are unfamiliar with working with satellite images I suggest you utilize the csv files. I will put out a tutorial on working with the satellite images in the near future. If you have any questions, please post them an I will try to answer them as soon as I can. If you have any questions related to the data, please refer to the data dictionary. Documentation: ------ Documentation for this dataset is an ongoing process given the complex and extensive process to preparing this dataset, amd the wide range of data sources. If you have a particular question please post it and I'll answer it as soon as possible. Upcoming Work: ------ As I mentioned above, I am going to use this data to build a machine learning model that will be able to predict the poverty for any region in any impoverished country! Stay tuned!