• Programming Problem 1: Use the ionosphere dataset in Matlab, and print your Matlab output.
  • How many elements are in this data set? How many values are used to describe each data element?

(i.e. what is the dimension of the feature vector for each data element?)

34,351

• Use the built-in k-nearest neighbors method function in Matlab to create a model to test new data.

Use k = 7 and the cosine distance metric.

• Find the k-fold loss, or the average fraction of testing elements that are misclassified when k-fold

cross- validation is used. Use the default value for the number of folds, i.e. 10.

• Find the predicted class for a new testing element, where each entry is chosen randomly and uniformly from [0,1].