Dynamic Evaluation of Intelligent Index System in Benchmark Regression: A Vertical and Horizontal Grading Approach

In the context of benchmark regression, the intelligent index system is measured using the entropy method. This approach involves objectively weighting each index by calculating the dispersion degree between data, which enables the retention of original data sample information to a great extent. The entropy method is commonly employed in evaluation models that involve data. However, it's not particularly suitable for panel data in time series as it lacks explanatory power. To avoid calculation deviation caused by static evaluation methods, we have opted for the vertical and horizontal grading method to evaluate the index system. This method maximizes the differences between evaluation objects in the time series three-dimensional data table and enables dynamic comparability of evaluation results. Please refer to the appendix for details on the steps involved in the vertical and horizontal grading method.