Network Support Data Analysis for Fault Identification Using Machine Learning

Network Support Data Analysis for Fault Identification Using Machine Learning

Shakila Basheer (King Khalid University, Abha, Saudi Arabia), Usha Devi Gandhi (VIT University, Vellore, India), Priyan M.K. (VIT University, Vellore, India) and Parthasarathy P. (VIT University, Vellore, India)
Copyright: © 2019 |Pages: 9
DOI: 10.4018/IJSI.2019040104
OnDemand PDF Download:
No Current Special Offers


Machine learning has gained immense popularity in a variety of fields as it has the ability to change the conventional workflow of a process. The abundance of data available serves as the motivation for this. This data can be exploited for a good deal of knowledge. In this article, we focus on operational data of networking devices that are deployed in different locations. This data can be used to predict faults in the devices. Usually, after the deployment of networking devices in customer site, troubleshooting these devices is difficult. Operational data of these devices is needed for this process. Manually analysing the machined produced operational data is tedious and complex due to enormity of data. Using machine learning techniques will be of greater help here as this will help automate the troubleshooting process, avoid human errors and save time for the technical solutions engineers.
Article Preview

2. Literature Survey

2.1. Introduction to Machine Learning

In an attempt to analyse the industrial data/machine produced data research fraternity has done a significant contribution. For industrial data analysis, a strong subject expertise is needed. But, the huge result sets and internal relationships between the workflow is sometimes beyond our subjective knowledge. To overcome this, a more generic framework for processing industrial data is needed. Mr. Mariusz Kamola, in his work (2015) has comes up with a defined set of rules for choosing the most required features for predictive analysis on industrial data. Clearly, the processing framework will differ depending on the use case and type of analysis. So, choice of a suitable Machine Learning algorithm is necessary.

Surya, Nithin, Prasanna, and Venkatesan (2016), gives a brief introduction to machine learning and discusses about various machine learning techniques and pre-processing techniques. The paper discuses about three main topics. They are:

  • Types of machine learning

  • Machine learning techniques

  • Linguistic pre-processing

Types of Machine Learning:

  • Supervised Learning: In this technique, knowledge is referred from training datasets. Example: classification and regression;

  • Unsupervised Learning: In unsupervised learning, there is no training datasets. In this technique, knowledge is inferred from input data that are not tagged. Example: clustering and dimensional Reduction;

  • Reinforcement Learning: A software agent is trained to make suitable decisions to be taken which will be based on the previous experience;

  • Machine Learning: Techniques discussed are, N-Gram and Markov Models, Neural Networks and Decision Tree classifiers;

  • Linguistic Pre-Processing: This step is a preparatory step which prepares the process to take place. This will ensure that the text will be in a form that would be understood by the machine. Here the context of a word is understood.

Complete Article List

Search this Journal:
Open Access Articles
Volume 10: 4 Issues (2022): Forthcoming, Available for Pre-Order
Volume 9: 4 Issues (2021): 3 Released, 1 Forthcoming
Volume 8: 4 Issues (2020)
Volume 7: 4 Issues (2019)
Volume 6: 4 Issues (2018)
Volume 5: 4 Issues (2017)
Volume 4: 4 Issues (2016)
Volume 3: 4 Issues (2015)
Volume 2: 4 Issues (2014)
Volume 1: 4 Issues (2013)
View Complete Journal Contents Listing