Question 31
UnclassifiedAn SVM with a non-linear RBF kernel works by:
Correct answer: B
Explanation
An RBF kernel lets an SVM handle non-linear data by using the kernel trick to compute similarities as if the inputs were mapped into a higher-dimensional feature space. In that space, a linear hyperplane can separate classes that are not linearly separable in the original input space.
Why each option is right or wrong
A. Fitting a linear hyperplane in the original input space
B. Implicitly mapping inputs into a higher-dimensional space where a linear separator exists
Under the kernel trick, the RBF kernel computes pairwise similarity as K(x, z) = exp(-\gamma\|x-z\|^2), which corresponds to an implicit feature mapping rather than an explicit coordinate transformation. In the induced feature space, the SVM solves the standard maximum-margin problem with a linear decision boundary, even though the original input space may be non-linearly separable.
C. Building decision trees over the support vectors
D. Applying PCA before fitting