Supervised Machine Learning with Python

Is a model parametric or not?

If you find yourself wondering whether or not a model is parametric, it's probably not the most important question to answer. You should select the modeling technique that best suits your data. However, a good rule of thumb is how many characteristics or parameters a model learns. If it's related to the feature space or dimensionality, it's probably parametric, for instance, learning the number of coefficients theta in a linear regression. If, instead, it's related to the number of samples, it's probably non-parametric, for instance, the depth of the decision tree or the number of neighbors in clustering.