The increase in electricity theft has become one of the main concerns of power distribution networks. Indeed, electricity theft could not only lead to financial losses, but also leads to reputation damage by reducing the quality of supply. With advanced sensing technologies of metering infrastructures, data collection of electricity consumption enables data-driven methods to emerge in such non-technical loss detections as an alternative to traditional experience-based human-centric approaches. In this context, such fraud prediction problems are generally a thematic of missing patterns, class imbalance, and higher level of cardinality where there are many possibilities that a single feature can assume. Therefore, this article is introduced specifically to solve data representation problem and increase the sparseness between different data classes. As a result, deeper representations than deep learning networks are introduced to repeatedly merge the learning models themselves into a more complex architecture in a sort of recurrent expansion. To verify the effectiveness of the proposed recurrent expansion of deep learning (REDL) approach, a realistic dataset of electricity theft is involved. Consequently, REDL has achieved excellent data mapping results proven by both visualization and numerical metrics and shows the ability of separating different classes with higher performance. Another important REDL feature of outliers correction has been also discovered in this study. Finally, comparison to some recent works also proved superiority of REDL model.