Predicting Heating/Cooling Loads with the Zoetrope Genetic Programming (ZGP) Versus Other Machine Learning Methods
摘要
This paper presents a comparison between a relatively new regression model called Zoetrope Genetic Programming (ZGP) and traditional machine learning techniques such as Linear Regression, Random Forest, Support Vector Classifier, and Multi Linear Perceptron. The application is a challenging heat load prediction problem with a real data set selected. The ZGP showed comparative results and was in second place for most of the metrics used. The Random forest still showed the best results. Analysis and justifications are shown in the rest of the paper.