Question24 In a soft-margin SVM, decreasing the slack penalty term C causes which of the following? (you can choose more than one) A smaller margin Less sensitivity to outliers More overfitting Less overfitting ResetMaximum marks: 1.5 Flag question undefined多项选择题
A
A smaller margin
B
Less sensitivity to outliers
C
More overfitting
D
Less overfitting

Question

Question24 In a soft-margin SVM, decreasing the slack penalty term C causes which of the following? (you can choose more than one) A smaller margin Less sensitivity to outliers More overfitting Less overfitting ResetMaximum marks: 1.5 Flag question undefined多项选择题

A

A smaller margin

B

Less sensitivity to outliers

C

More overfitting

D

Less overfitting

BlackTom AI · Accepted Answer

The question asks about the effect of decreasing the slack penalty term C in a soft-margin SVM, with multiple possible correct answers.
Option 1: 'A smaller margin' — This is not what happens when you decrease C. In soft-margin SVM, a smaller C increases regularization, which typically enlarges the margin to allow more slack. Therefore, a smaller margin is incorrect.
Option 2: 'Less sensitivity to outliers' — Decreasing C reduces the penalty for misclassifications, allowing the model to tolerate outliers more readily. This makes the decision boundary less influenced by individual outliers, so this statement is correct.
Option 3: 'More overfitting' — Lowering C reduces the model’s emphasis on fitting every training point (including outliers), which generally decreases overfitting. So this option is incorrect.
Option 4: 'Less overfitting' — As just noted, smaller C increases regularization and widens the margin, which tends to reduce overfitting. This option is correct.
In summary, decreasing C leads to a larger margin and greater tolerance for misclassified points, reducing overfitting and reducing sensitivity to outliers, while not producing a smaller margin or more overfitting.