Problem $3$

$(20$ pts$)$ Program: Code up a $k-$nearest$-$neighbor classifier. Your codes should include two files. One file is

the $kNN$ classifier function, which has the following four input arguments:

new: input variable for the test data set.

$x$ : input variable values for the training data set.

$y$ : output labels for the training data set.

$k$ : neighborhood size.

The output of the kNN classifier function should be the prediction results of the test data set. The other

code file should be the "main" file, which includes

Loading$/$Reading the training and test data sets.

Calling the kNN classifier function and return the prediction results. Specify three $k$ values: $k=5,$

$10,$ and $15.$

Figure $1$: Training Data Set

You should submit:

Three plots of the test data set with the predicted results colored as shown in Figure $1.$

Two files of codes. Don't submit the data sets.

The training and test data sets can be downloaded from Blackboard. You can use Python, MATLAB, or R$.$

Python is preferred. Moreover, you can put the two functions in a Jupyter Notebook.

Technical support can be also extended to MATLAB. However R is limited$.$

Reminder: Your final project will be based on Python, please start using Python as early as possible.

In the training data $("$training$\_$data.txt$"),$ the first two columns $($from left to right$)$ are input variables, the

last column is the output label. In the test data $("$test$\_$data.txt$"),$ all the two columns are input variables.