5.2. Fraud Detection In the context of fraud detection, the results obtained could be divided between the analysis of the classification models for detecting fraud in the acquired dataset and their interpretability (i.e., their capacity to provide insights into the way the algorithms identify fraud). The first type of analysis began with the application of a 10-fold cross-validation to compare the RF and XGB ability to identify fraud, this meant that the embedded dataset was split into subsets.
