Unique ID: WB28

Division: EDU
Issue Date: February 13th 2019
Last modified: February 22nd 2019
Collaborative

Targeting Poverty by Predicting Poverty: Using Machine Learning in Targeted Transfer Programs

Targeting Poverty by Predicting Poverty: Using Machine Learning in Targeted Transfer Programs

SDG: 01 - No Poverty

Targeting can be posed as a straightforward supervised learning problem, in the language of Machine Learning (ML). We will apply these techniques to see how we can improve on classification accuracy over existing programs. In this project we plan to use a wide array of data to attack this problem. We will ask: Are there better prediction functions that can be used on the existing survey platforms that would significantly improve targeting of poverty. A key to this proposal is the combination of two elements: machine learning tools and experimental methods. We will seek to use experiments (specifically at Give Directly and the historical experiment with Progressa) where individuals are randomly given transfers to independently and experimentally validate the efficacy of new targeting rules, both in targeting poverty and in targeting impact.

Project Sources
Project Sources
Type Of Institution: international organization
Region: Global
Country Area: Global
Id Country Regional: global
SDG Indicators
SDG Indicators
SDG: 01 - No Poverty
SDG Comments: 1.1, 1.2, 1.3, 1.4
Other
Other
Timeframe To Produce Indicator: NA
Write Your Own Review
You're reviewing:Targeting Poverty by Predicting Poverty: Using Machine Learning in Targeted Transfer Programs
Your Rating