Given the small training data, can you generalize well on the large testing set?

The goal here is to implement a learning algorithm that can generalize well on unseen data. You will be given 100 samples regarding sales data. The testing set consist of more than 19,000 samples.

Columns Available:

- Cost = Total Cost for Production.
- Ad Responses = Number of responses received per location regarding an advertisment.
- Location Participation = Rate of location participation to a certain corporate objective.