Aggregate density-based concept drift identification for dynamic sensor data models.

DATA modeling; BUSINESS communication; DECISION making; INTERNET of things; CONCEPTS

The reduced costs of embedded systems and sensor technology coupled with the increased speed in communication enables businesses and consumers to deploy a large number of sensing devices. This conjunction of technologies has come to be known as the Internet of Things (IoT). Data collected from IoT devices are continuously increasing, and many approaches have been proposed to deal with the big data that is now generated. Multiple artificial intelligent techniques have been proposed and used to extract knowledge out of these continuously growing datasets. In this paper, we demonstrate that a better understanding of data can be achieved through dynamic modeling. This dynamic behavior is observed in many practical scenarios and needs to be taken into account to have a higher accuracy in prediction and analysis for policy making and business-related decisions. We propose and test a novel methodology to detect the dynamic nature of data over time. Machine learning models have been known to suffer from changes in streaming data over time which is defined as concept drift and therefore by detecting this phenomena such models can be improved. [ABSTRACT FROM AUTHOR]

Copyright of Neural Computing & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)