Predicting Adult Income Utilizing Various Artificial Intelligence Models

Abstract

This study examines the growth of artificial intelligence (AI) models to forecast adult yearly income, with an objective to use new computational tools to gain a better scientific perspective of economic dynamics. To this end, the study makes a contribution to the development of more effective labour market strategies. Using AI algorithms, the study handles economic information from the UCI Adult Dataset. The performance of four machine learning (ML) models was studied which included support vector machines (SVM), K nearest neighbors (KNN), random forests (RF) and Logistic regression (LR). Top eight predictive features were found using the Recursive Feature Elimination (RFE) approach before the implementation of these models. Among the tested models, SVM was the best performing producing an accuracy of 82.6%, recall of 88.9% and hence the most effective at predicting income level. The results stress the possible role of artificial intelligence in doing financial data analysis and predicting revenue, focusing on its use in employment policies, financial planning, and economic research work.