Python Machine Learning By Example - Second Edition

Choosing between linear and RBF kernels

Of course, linear separability is the rule of thumb when choosing the right kernel to start with. However, most of the time, this is very difficult to identify, unless you have sufficient prior knowledge of the dataset, or its features are of low dimensions (1 to 3).

Some general prior knowledge we have include: text data is often linearly separable, while data generated from the XOR function is not.

Now, let's look at the following three scenarios where linear kernel is favored over RBF:

Scenario 1: Both the numbers of features and instances are large (more than 10⁴ or 10⁵). Since the dimension of the feature space is high enough, additional features as a result of RBF transformation will not provide ...