6. As we decrease k in the k-nearest neighbor algorithm, it is possible for RMSE (validation) to go up.

True or False

7. Suppose we have a 100 observations for variables X and Y and we implement k-nearest neighbors. A fit based on k = 90 is a more flexible fit than k = 3?

True or False

10. With small data sets, nonparametric approaches tend to produce more accurate estimates of the underlying model than parametric models.

True or False