Can the WEKA Data Mining Tool be Used in Developing an Economic Growth Model?

Abstract

WEKA is a free and open source software program that is written in JAVA™ language and contains GUI for interacting with data files and produces visual results. Although conventional economic theories assert that inflation and economic growth are negatively related, researchers found throughout the last 5 decades using different  models and datasets, in favor and against that conventional beliefs. To help predict the expected economic growth or Inflation using 7 attributes such as: war, totch, pssurp, drelief, inflation, invsh, and growth found in the World Bank  dataset (link was provided above), three models/algorithms and 2 testing options were used to find out the best model for decision support. As far as the Correlation Coefficient or measurement of how well the predicted values from a forecast model "fit" with the real-life data is concerned, Linear regression algorithm seems to predict the expected growth much better than other two algorithms in this study namely ZeroR and REPTree using both 10 fold cross validation as well as 75 percentage split test options.