This study aimed to analyze credit card data and improve the detection of frauds carried out with this type of payment through the construction of predictive models using machine learning. To achieve this, an exploratory analysis was conducted to understand the dataset and extract insights, as well as to verify possible correlations among the attributes. Then, two predictive models were created using machine learning, employing Linear Regression and Decision Tree algorithms. Finally, the models' results were evaluated to identify the one with the best performance based on the metrics of Recall, AUC, and confusion matrix.
The results obtained with the best machine learning model were condensed into images to generate a straightforward presentation of what the company could achieve with the use of this algorithm.
