Every chapter has exercises. One example from Ch. 14 - Kernels is this: > Exercise 14.2 Linear separability > (Source: Koller..) Consider fitting an SVM with

Every chapter has exercises. One example from Ch. 14 - Kernels is this: > Exercise 14.2 Linear separability > (Source: Koller..) Consider fitting an SVM with C > 0 to a dataset that is linearly separable. Is the resulting decision boundary guaranteed to separate the classes? etc. Many exercises are proofs or derivations, and the book is full of (algorithm/optimization) defining/bounds approximation/ otherwise pragmatic information.