From "Reinforcement Learning: An Introduction":

Radial basis functions (RBFs) are the natural generalization of coarse coding to continuous-valued features. Rather than each feature being either 0 or 1, it can be anything in the interval , reflecting various degrees to which the feature is present. A typical RBF feature has a gaussian (bell-shaped) response, dependent only on the distance between the state, s, and the feature's prototypical or center state and relative to the feature's width: $ \Phi_s(i) = \exp \left( -\frac{||s-c_i||^2}{2\sigma_i^2} \right) $...

In addition, some learning methods for RBF networks change the centers and widths of the features as well.

See also: https://www.math.ucdavis.edu/~saito/data/jim/buhmann-actanumerica.pdf

From sklearn: "For instance many elements used in the objective function of a learning algorithm (such as the RBF kernel of Support Vector Machines or the L1 and L2 regularizers of linear models) assume that **all features are centered around 0 and have variance in the same order**. If a feature has a variance that is orders of magnitude larger that others, it might dominate the objective function and make the estimator unable to learn from other features correctly as expected."

Community content is available under CC-BY-SA
unless otherwise noted.