Prediction of purchasing power of Google store based on deep ensemble learning model

Abstract

Aiming at the defect that the ensemble learning model such as Light Gradient Boosting Machine only mines the data information once, which can not automatically refine the granularity of data mining and dig into the more potential internal correlation information of data, the ensemble learning model is made into a deep form by sliding window and deepening, and the deep ensemble learning is proposed. Sliding window enables the ensemble learning model to automatically refine the granularity of data mining, so as to dig deeper into the potential internal correlation information in the data, and at the same time endue it with certain representation learning ability. Based on the sliding window, the deepening step further improves the representation learning ability of the model. Finally, the results show that the prediction accuracy of the deep ensemble learning model is 6.16 percentage points higher than that of the original ensemble learning model.