Research on collaborative status evaluation methods for Yangtze River Protection Projects based on random forest

Abstract

The construction system of the Yangtze River Protection Project is relatively complex, and its synergistic state is difficult to be evaluated directly through the engineering construction information, and the existing method rely on the subjective evaluation of the participants. Based on the principle of servitude, a random forest-based evaluation method of the project construction synergistic state was proposed through the fast variable inversion of the sequential parameter. Through engineering research and literature analysis, we select two ordinal covariates, namely, the degree of interest satisfaction and the degree of information matching, and seven fast variables, such as the number of construction reworks, to construct the quantitative relationship between the ordinal covariates and the fast variables, and then carry out statistical analysis. Then, the training set and test set are constructed to establish the prediction model of the two ordinal covariates of interest and information, and the model is verified and evaluated by the ten-fold cross-validation method and the test set verification, respectively, and the matching rule between the ordinal covariates and the collaborative state is established as the basis of evaluation. This matching relationship between ordinal parameters and synergistic state matches the predicted values of ordinal parameters with the synergistic state of subsystems, and obtains the synergistic state of the corresponding subsystems, which provides a basis for the evaluation of the synergistic state of engineering construction.

Keywords

Yangtze River Protection Project / collaborative evaluation / order parameters / random forest