Experience Estimation of Edge Cloud Network based on the Prediction of Data Volume

Statistical learning (SL), represented by machine learning (ML) and deep learning (DL), has become a research hotspot in the recent years due to its excellent performance. However, SL-enabled network models are data-driven and enough samples are required to ensure their performance. With the development of edge cloud and cloud computing, data technology, information technology and communication technology (DICT) has begun to expand into the field of industrial customers, and data acquisition is a time-consuming and costly task in relevant industries. Therefore, how much data is needed to build a statistical learning model is the focus of the industry, which directly determines whether to promote a project. Motivated by this, this paper proposes a data volume prediction scheme of customer experience in edge cloud network, where the model training with good generalization performance can be determined by the data spectrum (Fourier transform calculation). Our experimental results show that the correlation between the proposed data spectrum and the performance of the model has been improved by an average of 91% compared with the simple data size, and the simpler the model is, the better the performance will be.