Predicting Life Satisfaction with XGBoost
In this project, we build a machine learning model that can be used to classify individuals life satisfaction using variables in the National Income Dynamic Studies (NIDS) database.
NIDS is a face-to-face longitudinal survey of individuals and households living in South Africa and is the first 'nationally representative' panel study that strives to document various changes over time in the income, expenditures, assets, access to services, education, health, agriculture and other dimensions of well-being of 11 895 households
Specifically, we employ the xgboost model (using the eXtreme Gradient Boosting package in R). The variables chosen are strongly captured by the survey and thus contain few missing values. This model is purely a trial run, and it may be the case that alternative ML algorithms and/or different data processing techniques that will yield more accurate results