Abstract:With the development of 5G technology and Internet of Things, all kinds of real life data are collected and recorded by a large number of sensors. It is of great significance to mine and analyze the hidden information in the data for applications like future prediction. However, due to interferences or instability of collection equipment, collected sensory data are often incomplete, and this incompleteness hinders the in-depth analysis of data in the cloud. Therefore, processing around missing values is significant. Relying on cloud machine learning methods is not enough to deal with the problem of missing data in the Artificial Intelligence of Things (AIoT) environment, however, edge computing provides a promising solution. In this article, gated recurrent units filling is employed at the edge nodes. A mobile edge node can not only find the historical information of the current missing data node but also acquire the data of the nodes adjacent to the missing data node. These ensure that the missing data are restored to the maximum extent at the source. The experimental results show that the missing value filling based on edge computing not only outperforms other filling methods in quality but also greatly reduces the energy consumption in AIoT.

Missing Value Filling Based on the Collaboration of Cloud and Edge in Artificial Intelligence of Things