video1.0<iframe src="https://www.loom.com/embed/01be0d9c78f849d2bf437b6214015f63" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/01be0d9c78f849d2bf437b6214015f63-ff79b966a1e33ae1.gif553.791In this video, I walk through my process of conducting exploratory data analysis on a dataset with over 5,000 unique consumer IDs and a fraud ratio of 6.5%. I emphasize the importance of understanding data distributions and the need for feature engineering to identify signals related to fraud. I also highlight key findings, such as the behavior differences between fraudulent and non-fraudulent users. Please review the insights and consider how we can apply these findings to improve our model.