KNN: Get optimized parameters...

Perform an automated search for the optimal KNN parameter settings.


Evaluation method
The method to be used for estimating the classification accuracy, which in turn is used to guide the model search. Supported methods are 10-fold cross-validation and leave-one-out.
k max
The maximum value of the parameter k.
Number of seeds
The size of the parameter neighbourhood to be searched.
Learning rate
The rate at which the parameter neighbourhood is shrunk (per iteration).

See also:

kNN classifiers 1.1.2. Model selection
kNN classifiers 1.1. Improving classification accuracy
kNN classifiers 1. What is a kNN classifier?
kNN classifiers

Links to this page

© Ola Söder, July 18, 2008