This solution predicts the Loss due to default of the borrowers who are most likely to default on their Consumer loans in Peer-to-Peer lending. During the training stage, the solution understands the dataset, handles missing data and class imbalance, conducts feature interaction on the training data and selects a subset of best features based on feature importance. It then trains on multiple ML models, identifies the best performing model and tunes it accordingly. This trained model is then selected for prediction on the test data.