Exploiting Linear Support Vector Machine for Correlation-Based High Dimensional Data Classification in Wireless Sensor Networks

Linear Support Vector Machine (LSVM) has proven to be an effective approach for link classification in sensor networks. In this paper, we present a data-driven framework for reliable link classification that models Kernelized Linear Support Vector Machine (KLSVM) to produce stable and consistent results. KLSVM is a linear classifying technique that learns the “best” parameter settings. We investigated its application to model and capture two phenomena: High dimensional multi-category classification and Spatiotemporal data correlation in wireless sensor network (WSN). In addition, the technique also detects anomalies within the network. With the optimized selection of the linear kernel hyperparameters, the technique models high-dimensional data classification and the examined packet traces exhibit correlations between link features. Link features with Packet Reception Rate (PRR) greater than 50% show a high degree of negative correlation while the other sensor node observations show a moderate degree of positive correlation. The model gives a good visual intuition of the network behavior. The efficiency of the supervised learning technique is studied over real dataset obtained from a WSN testbed. To achieve that, we examined packet traces from the 802.15.4 network. The technique has a good performance on link quality estimation accuracy and a precise anomaly detection of sensor nodes within the network.

Download Full-text

A Hybrid Feature Selection Method Based on Symmetrical Uncertainty and Support Vector Machine for High-Dimensional Data Classification

Intelligent Information and Database Systems - Lecture Notes in Computer Science ◽

10.1007/978-3-319-54472-4_67 ◽

2017 ◽

pp. 721-727 ◽

Cited By ~ 2

Author(s):

Yongjun Piao ◽

Keun Ho Ryu

Keyword(s):

Support Vector Machine ◽

Feature Selection ◽

High Dimensional Data ◽

Feature Selection Method ◽

Data Classification ◽

Selection Method ◽

High Dimensional ◽

Support Vector ◽

Symmetrical Uncertainty

Download Full-text

Fast outlier detection for high-dimensional data of wireless sensor networks

International Journal of Distributed Sensor Networks ◽

10.1177/1550147720963835 ◽

2020 ◽

Vol 16 (10) ◽

pp. 155014772096383

Author(s):

Yan Qiao ◽

Xinhong Cui ◽

Peng Jin ◽

Wu Zhang

Keyword(s):

Wireless Sensor Networks ◽

Support Vector Machine ◽

Sensor Networks ◽

Outlier Detection ◽

Large Scale ◽

Deep Belief Network ◽

Wireless Sensor ◽

High Dimensional ◽

Support Vector ◽

Belief Network

This article addresses the problem of outlier detection for wireless sensor networks. As increasing amounts of observational data are tending to be high-dimensional and large scale, it is becoming increasingly difficult for existing techniques to perform outlier detection accurately and efficiently. Although dimensionality reduction tools (such as deep belief network) have been utilized to compress the high-dimensional data to support outlier detection, these methods may not achieve the desired performance due to the special distribution of the compressed data. Furthermore, because most existed classification methods must solve a quadratic optimization problem in their training stage, they cannot perform well in large-scale datasets. In this article, we developed a new form of classification model called “deep belief network online quarter-sphere support vector machine,” which combines deep belief network with online quarter-sphere one-class support vector machine. Based on this model, we first propose a model training method that learns the radius of the quarter sphere by a sorting method. Then, an online testing method is proposed to perform online outlier detection without supervision. Finally, we compare the proposed method with the state of the arts using extensive experiments. The experimental results show that our method not only reduces the computational cost by three orders of magnitude but also improves the detection accuracy by 3%–5%.

Download Full-text