UPLIFT: Usage of Poverty Level Indicators to Facilitate Transformation
Namrata Ramesh, Diya Chandra, Chiranth Jawahar, Gowri Srinivasa · 2021 IEEE Bombay Section Signature Conference (IBSSC) · 2021
UPLIFT: Usage of Poverty Level Indicators to Facilitate Transformation, as the name suggests, is a study that aims at analysing data consisting of various socio-economic factors that affect households in order to alleviate poverty by understanding and identifying various aspects that may have an impact on poverty levels. As a use case, we have studied data pertaining to the Latin American nation of Costa Rica. Initially, we perform exploratory data analysis to discover patterns in the data and gain an insight to the relationship between attributes. We then perform feature selection and normalization to render the data suitable for modeling with machine learning techniques. We use these models to predict the economic health of each household, mapping them into four categories corresponding to the level of poverty. Finally, the data is reinterpreted in the light of this classification to elicit the factors that strongly affect each category, towards recommending methods for improvement and an appropriate allocation of social welfare benefits.