Title: Outlier detection of time series with a novel hybrid method in cloud computing

Authors: Qi Liu; Zhen Wang; Xiaodong Liu; Nigel Linge

Addresses: Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science and Technology, Nanjing, 210044, China; School of Computing, Edinburgh Napier University, Edinburgh, Scotland, UK ' School of Computer and Software, Nanjing University of Information Science and Technology, Nanjing, China ' School of Computing, Edinburgh Napier University, 10 Colinton Road, Edinburgh EH10 5DT, UK ' School of Computing, Science and Engineering, The University of Salford, Salford, Greater Manchester, M5 4WT, UK

Abstract: In the wake of the developments in science and technology, cloud computing has obtained more attention in different fields. Meanwhile, outlier detection for data mining in cloud computing is playing significant role in different research domains and massive research works have been devoted to outlier detection. However, the existing available methods require lengthy computation time. Therefore, the improved algorithm of outlier detection, which has higher performance to detect outliers, is presented. In this paper, the proposed method, which is an improved spectral clustering algorithm (SKM++), is fit for handling outliers. Then, pruning data can reduce computational complexity and combine distance-based method Manhattan distance (distm) to obtain outlier score. Finally, the method confirms the outlier by extreme analysis. This paper validates the presented method by experiments with real collected data by sensors and comparison against the existing approaches. The experimental results show that our proposed method outperforms the existing.

Keywords: cloud computing; data mining; outlier detection; spectral clustering; Manhattan distance.

DOI: 10.1504/IJHPCN.2019.102350

International Journal of High Performance Computing and Networking, 2019 Vol.14 No.4, pp.435 - 443

Received: 05 Sep 2017
Accepted: 21 Mar 2018

Published online: 19 Sep 2019 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article