Appendix_B_Exploratory_data_analysis_SOM_results

<p dir="ltr">Contains the output of the self-organizing maps for the different types of faults for both the cost and node signature matrix data. Note that the colour legend accompanying the output of each self-organising map represents how closely the weights of a neuron correspond to the weights of neighbouring neurons. Darker colours indicate that the neuron’s weight differs greatly from the weights of neighbouring neurons, and lighter colours indicate that the neuron’s weight more closely resembles the weights of neighbouring neurons.</p>