Adult Income Classification using Machine Learning Techniques

Abstract

Nowadays, the economic growth of country is often measured by the increasing gross domestic product (GDP). GDP is an effective indicator in the national economic accounting system and has a significant reference function for political decision and regional development. GDP improvement shows economic development level, national income and spending power of a country or region. In actual accounting, three methods can be used to compute GDP namely production, expenditure, or income approach which respectively reflect gross domestic product and its composition from different aspects. This paper is presented to predict the GDP of person based on adult income data with the age from 17 to 90 using popular supervised learning techniques. The models are tested on more than 30000 data records of over forty countries especially United States. The experiments show that the different results between Naïve Bayes, J48 and Random Forest classifiers.