Optimal Weighted Average Prediction and Correction in Big Sensor Data Using Fruit Fly Algorithm and Support Vector Machine

Some techniques have been developed in recent years for processing sensor data on cloud, such as sensor-cloud. However, these techniques do not provide efficient support on fast detection and location of errors in big sensor data sets. In this paper, a novel data error detection approach which exploits the full computational potential of cloud platform and the network feature of Wireless Sensor Network (WSN) for fast data error detection in the big sensor datasets is developed. In the proposed methodology, the input big sensor data set is initially preprocessed by dimensionality reduction for better prediction. Then the features of errors in the big sensor datasets are calculated using Optimal Weighted Average Prediction (OWAP). Then the optimization weights are selected using Fruit Fly Algorithm. The computational power and scalability of the Cloud environment are fully exploited for supporting the real time fast error detection of the big sensor data sets. Support Vector Machine (SVM) aids in classifying the errors. The errors are predefined and the clustering for the whole network is carried out with Kernel Based Fuzzy C-means clustering (KFCM). Finally, the errors are localized.