1. create a predictive model based on historical claim data after cleaning the data. Team is concerned about the fraud detection accuracy as well as the key drivers that cause fraudulence. Tasked with identifying first-party physical damage fraudulence and explaining the indicators of fraudulent claims.
2. Models to be evaluated using F1 score. The submission file should be a CSV containing two columns: claim_nbr and prediction (the predicted fraud indicator (0 or 1), not the probability)