Study, Then Answer — easy Machine Learning in Code problem | Incognition

Study, Then Answer

10 pts · 10 coins

Machine Learning in CodeEasy

Study, Then Answer

A market stall sorts fruit by two measurements: weight in grams and a sweetness score.

The owner has a few fruits that are already labelled. Now some new fruits arrive with no label.

Task: write predict_fruit(train_rows, train_labels, new_rows).

train_rows is a list of rows, one inner list of numbers per fruit. train_labels holds the answer for each of those rows, in the same order.
new_rows is a list of rows shaped the same way, but with no answers.
Make a KNeighborsClassifier with n_neighbors=3, let it learn from the labelled fruits, then ask it about the new ones.
Return the predicted labels as a plain Python list, one per new row, in the same order as new_rows.

Examples

train_rows = [[150, 7], [170, 8], [130, 3], [110, 2], [160, 6]]

train_labels = ["apple", "apple", "lemon", "lemon", "apple"]

new_rows = [[155, 7], [120, 3]]

returns ['apple', 'lemon']

train_rows = [[4.1, 30], [3.8, 28], [9.5, 61], [10.2, 66], [4.4, 33], [9.9, 58]]

train_labels = ["cat", "cat", "dog", "dog", "cat", "dog"]

new_rows = [[9.0, 60], [4.0, 31], [10.5, 70]]

returns ['dog', 'cat', 'dog']

Hints

solution.py

1from sklearn.neighbors import KNeighborsClassifier
2
3
4def predict_fruit(train_rows, train_labels, new_rows):
5    pass
6