Volume 18, No. 6, 2021
CID: Central Tendency Based Noise Removal Technique For Iot Data
V. A. Jane , Dr. L. Arockiam
Abstract
The Internet of Things (IoT) is a crucial technique that enables well-organized and reliable solutions for various areas' development. Agriculture is one of the most worried IoT areas, with IoT-based solutions being used to automate the management and evaluation system with the least amount of human participation. Every second, a large-scale IoT-based agricultural ecosystem creates a tremendous quantity of data. The agro-production ecosystem is complicated, and there are several inconsistencies in the raw data that assessment and mining cannot be directly tracked. This research presents a strategy called Detection and Removal of Noise (CID) to deal with these anomalies in IoT agriculture data. Utilizing measurements of central tendency, the suggested approach eliminates null values, incorrect values, repeating values, unfinished values, and inappropriate values. In addition, a comparison of current noise reduction strategies was carried out, and the effectiveness was assessed using the Support Vector Machine (SVM) classification. To improve accuracy of classification, noisy data is removed in this suggested investigation. The CID approach will help improve the quality of data obtained in agricultural settings.
Pages: 2210-2217
Keywords: Noise, Data cleaning, IoT, Pre processing, Noise removal, Smart Agriculture