Abstract: This paper presents a computational implementation of a Probabilistic Neural Network for obtaining patterns of partial discharges in power cables XLPE. The experimental measurement data of the power cables are obtained in the Laboratory Testing Equipment and Materials (LAPEM), which is a certified laboratory property of the Comisión Federal de Electricidad (CFE) in México. These data implicitly contain the patterns of partial discharges and are used to carry out the training and testing of Probabilistic Neural Network. In order to illustrate the reliability and validity of the proposed computational implementation, the results obtained by this proposed implementation are compared with those calculated by the methodologies given in the standard IEC 60270.