Nowadays, advances in sensing and information processing
technologies have made possible to remotely asses the power
grid assets condition. However, this involves continuous operation,
which produces a huge amount of streaming sensory
data that requires analysis. To address this, machine learning
techniques have been successfully used to discover features
and specific patterns that can differentiate between partial
discharges and noise. However, labelled data is required in
order to build predictive models for PD monitoring, such data