SVM Quiz
Check your understanding of support vector machines, which points define the boundary, what the kernel trick avoids computing, and how C and γ shape the fit.
Exercise
Which training points actually determine an SVM's decision boundary?
Need a hint?
Most points could be deleted without changing the solution.
Exercise
Maximizing the margin is equivalent to minimizing what?
Need a hint?
The margin width is 2/‖w‖.
Exercise
What does the kernel trick let an SVM do?
Need a hint?
K(x, x') = φ(x)·φ(x') — but you never compute φ.
Exercise
What does a very large γ in an RBF kernel do to the decision boundary?
Need a hint?
γ controls how fast similarity decays with distance.
Exercise
In a soft-margin SVM, what does a very large C encourage?
Need a hint?
C is the price of each margin violation.
Exercise
Swapping a linear SVM for an RBF-kernel SVM (to fit curved boundaries) changes which slot of the ML project loop?
Need a hint?
The kernel changes the family of boundaries the model can represent.
Exercise
Transfer test — a technique this course did NOT teach: SMOTE synthesizes new minority-class examples to fix class imbalance. Which slot does it modify?
Need a hint?
It edits the training distribution the model learns from.