Threshold tuning in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to set the threshold for binary classification.

ML Python

threshold = [1]

Drag options to blanks, or click blank then click option'

A0

B1.0

C0.5

D-0.5

Attempts:

3 left

💡 Hint

Common Mistakes

Using threshold values outside the 0-1 range.

Setting threshold to 0 or 1 which makes classification trivial.

The typical threshold for binary classification is 0.5, meaning predictions above 0.5 are positive.

2fill in blank

medium

Complete the code to convert probabilities to binary predictions using the threshold.

ML Python

predictions = (probabilities > [1]).astype(int)

Drag options to blanks, or click blank then click option'

A1.0

B0.5

C0.7

D0.3

Attempts:

3 left

💡 Hint

Common Mistakes

Using threshold greater than 1 or less than 0.

Not converting boolean array to integers.

Using 0.5 as threshold converts probabilities above 0.5 to class 1, else 0.

3fill in blank

hard

Fix the error in the code to compute true positives using the threshold.

ML Python

true_positives = ((y_true == 1) & (y_scores > [1])).sum()

Drag options to blanks, or click blank then click option'

A0

B-1

C1

D0.5

Attempts:

3 left

💡 Hint

Common Mistakes

Using threshold 0 or 1 which may not capture positives correctly.

Using negative threshold values.

The threshold 0.5 is used to decide positive predictions from scores.

4fill in blank

hard

Fill both blanks to create a dictionary of thresholds and their corresponding F1 scores.

ML Python

f1_scores = {thr: f1_score(y_true, (y_scores > [1]).astype(int)) for thr in [2]

Drag options to blanks, or click blank then click option'

Athr

B[0.1, 0.3, 0.5, 0.7, 0.9]

Cthresholds

Dy_scores

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong variable names for threshold or list.

Not converting boolean to int before computing F1 score.

Use 'thr' as the threshold variable and a list of threshold values to compute F1 scores.

5fill in blank

hard

Fill all three blanks to select the best threshold based on maximum F1 score.

ML Python

best_threshold = max(f1_scores, key=lambda [1]: f1_scores[[2]])
print(f"Best threshold: [3]")

Drag options to blanks, or click blank then click option'

Athr

Cbest_threshold

Dthreshold

Attempts:

3 left

💡 Hint

Common Mistakes

Using different variable names inside lambda and key lookup.

Printing a variable that is not defined.

Use 'thr' as the lambda argument and key to find max, then print best_threshold.