Investigation of the Effective Use of Ensemble Learning Algorithms for Cyber Data Analytics –The Prediction of the Customer Revenue on the Google Merchandise Store (GStore)

Generally, through research and development (R&D), cyber technology brings significant innovation to the world due to the repaid development of the cyber-physical systems (CPS). A large amount of data can be connected from the edge (e.g., local server, device, and sensor) to the cloud system. The magnitude of data in the CPS has dramatically increased from the terabyte, petabyte to exabyte. Moreover, cyber data analytics has attracted much attention from researchers with various applications, including e-commerce. Also, data analytics is presently used in such fields as retailer and marketing, so exploring the related data to obtain useful information may enhance enterprise revenue. Further, the ensemble learning algorithms are the new type of machine learning algorithms widely applied in a different field. The ensemble learning algorithms are data analytics technologies, which combine multiple machine learning models to improve computation performance. For the e-commerce field, however, how to analyze a large amount of data is challenging for enterprises. Thus, through ensemble learning algorithms, we used to analyze the Google Merchandise Store (GStore) data. The results are a reference to marketing decision making for an enterprise.