| dc.description.abstract | An important aspect of communication is involved in its cellular network. To meet the demands, communication requires the next generation cellular network, i.e., self organizing networks (SON). In order to implement a self-organizing network, its subsections have to be known and optimized using certain rules. The objective of this document is to deal with one of the subsections called “Self-healing: Fault identification,” in particular by conducting analysis on the Telstra cellular network and predicting its disruptions. First, the prediction of the disruptions can be determined by establishing the machine learning algorithms upon Telstra data. Thus, the classification of faults could be used for finding the nature of the disruptions. Because the appropriate algorithm is chosen by the trial-and-error method, there is no one particular algorithm that fits particular data. Thus, data has to be pre-processed for the algorithms to be applied. Here, the Python Sci-kit module was used as a tool for developing the predictive model. As a note, there are many other tools like R, MATLAB, Rattle, KNIME, etc. that can be used for machine learning. Then, the nature of the faults was identified and investigated to drive customer advocacy. | en_US |
