Integrated Learning-Based Feature Extraction and Regression Analysis

Integrated learning refers to the method of statistically deriving new results from predictions made by some weak regressors, which can greatly improve the appropriateness and reasonableness of predictions and fits. In this paper, integrated learning algorithms are applied to the study of industrial wastewater impact factors for the first time. Here, the already mature random forest regression method based on Bagging method for integration of decision tree regressors for feature extraction and AdaBoost integration method for model fitting are mainly applied.