The Least Energy Criterion LEC and Its Application in Model Selection

摘要

This paper considers the statistic LEC and its application on econometric model selection. The idea is to apply the concept of least energy principle, and based on information entropy propose the least energy criterion expression LEC. It is shown that using the idea of comparing the LEC of model residual sequences and selecting the least LEC one as the real model is feasible. For practice, this paper takes the data of China export volume from 1978 to 2011 as an example to verify the idea. The conclusion shows that using least energy criterion to select model is effective.