Diagnosis of Modern Economy Using Repetitive Identification Approach for the Prediction of GDP Per Capita

System identification for economic health monitoring in modern economy is moving to the forefront of worldwide research activities. Such monitoring for economic development and management is important for the public welfare related to technological breakthrough such as the Industrial Revolution, leading to new construction technology. In economy practice, however, there are many situations in which a feedback identification system is given model uncertainties and uncertainty of measurement. Aiming for accurate economy model updating, this paper presents the diagnosis of the GDP per capita trend through an automatic repetitive sifting process. It shows how a statistical confidence interval based model updating approach can be applied to the health evaluation of economic development via prediction of GDP per capita over time. The model updating approach uses the confidence interval of the estimated economic parameters to determine their statistical significance in order to mitigate the model uncertainties and measurement errors. If the parameters’ confidence interval covers the “null” value, it is statistically sustainable to truncate such parameters. The remaining parameters will repetitively undergo such parameter sifting process until all the parameters’ statistical significance cannot be further improved. Consequently, the proposed repetitive identification approach promotes the accuracy of the prediction of GDP per capita, assisting the assessment of modern economic trend.