Use the training data to estimate the model(s). After that, use the validation data to assess your model's performance and to see whether you have overfitting. Use the representative data to analyze the implications of your model and to figure out what would happen if you implemented a specific plan. For steps 1 and 3 below, explain what models you used. For these steps, you will have to explain what model(s) you ran with the training sample and how they did in the validation sample. Then, you will have to create new variables in the representative data that tell us what offer each person in the representative data should get. For step 2, explain what percentage of customers in the representative data would get each message type. For steps 4 - 8, focus just on the representative data