This document discusses using artificial neural networks (ANNs) and statistical techniques to classify partial discharge (PD) defects within cross-linked polyethylene (XLPE) medium voltage cables. PD measurements were taken from six cables with different defects and voltages. Statistical features were extracted from the 3D PD patterns to form the input for various ANNs for classification. 72 different ANN structures were analyzed to determine the most effective and optimal classification technique, based on metrics like mean square error and accuracy. The proposed approach achieved high recognition rates for identifying different types of PD defects within XLPE cables.