A classification model was developed to predict if a photo of the Tokyo Tower was taken in

the day or in the night, based on the mean of the saturation values.

A visual representation of results from testing this model on $21$ photos with the tag 'Tokyo

Tower' is shown below. Each of the photos were sourced from the website Unsplash and

were labelled as actually being taken in the day or the night.

Use the plot above to complete the blanks in the statements below. Be careful to not type

any blank spaces, to type the names of the variables and$/$or levels exactly as they are given

in the visualisation or instruction, and to give any percentages rounded to one decimal

place $($e$.$g$.25\mathrm{.}5\%$ or $32\mathrm{.}0\%).$

The decision rule shown above $($which may not be a good$/$sensible one$)$ can be described as:

If

is less than

$,$ classify the photo as

$,$ else classify the photo as

The PCC for the classification model is

The baseline model for this data would be to always classify the photo as