Search code examples
machine-learningknn

What are the discrete and continuous hyper-parameters for KNN?

I have this table

It says "Example classifiers in Scikit-Learn and their hyper-parameters. Generally, hyper-parameters can be (a) discrete, e.g., number of neighbors in kNN, or (b) continuous. e.g., the value of penalty in logistic regression."

enter image description here

as you can see, KNN has 3 hyper-param.

2 discrete & 1 continuous OK,

I know that K in KNN is one hyper-parameter?

So What are the other discrete and continuous hyper-param they are talking about?

Solution

  • I found the answer

    I think there is mistake in that research

    as the hyper-param are 2 continuous and 1 discrete

    hyper-param

    1- N neighbors (continuous)

    2- Weights (cont)

    3- Leaf Size (discrete)