Now let's use KNN for a regression task. Let's use a person's height and sex to predict their weight. The following tables list our training and testing sets:
|
Height |
Sex |
Weight |
|
158 cm |
male |
64 kg |
|
170 cm |
male |
66 kg |
|
183 cm |
male |
84 kg |
|
191 cm |
male |
80 kg |
|
155 cm |
female |
49 kg |
|
163 cm |
female |
59 kg |
|
180 cm |
female |
67 kg |
|
158 cm |
female |
54 kg |
|
178 cm |
female |
77 kg |
|
Height |
Sex |
Weight |
|
168 cm |
male |
65 kg |
|
170 cm |
male |
61 kg |
|
160 cm |
female |
52 kg |
|
169 cm |
female |
67 kg |
We will instantiate and fit KNeighborsRegressor, and use it to predict weights. In this dataset, sex has already been coded as a binary-valued feature. Notice that this feature ranges from 0 to 1, while the values of the feature representing the person's height range from 155 to...