Throughout the previous examples you have seen how the k-Nearest Neighbour method can be used to solve practical machine learning problems. While the customers are happy, you might still feel like there are unanswered questions related to the problems. Was k-NN really the best method to use here? What about those unused features in the dataset? Could the model be improved if they were included as well? These are some of the questions that are useful to think about when working with a machine learning problem and that, in the end, can lead to improved models. Upcoming sections of this course will give you the tools to reason about and answer these questions.