knn_cv function

Train a k nearest neighbors (knn) classifer via cross validation (cv).

Train a k nearest neighbors (knn) classifer via cross validation (cv). The number of folds and the set of the number of neihbors to consider may be specified.

knn_cv(xy, k.cv = 5, kvec = seq(1, 47, by = 2))

Arguments

• xy: Data frame with the data matrix x as the first set of columns and the vector y as the last column.
• k.cv: scalar. number of folds to use. default is 5.
• kvec: vector. set of neighbors to consider. default is odd integers between 1 and 47 (inclusive).

Returns

• kvec: set of neighbors considered

• error: vector of misclassification error rates corresponding to kvec

• k.best: number of neighbors with lowest error rate

• k.cv: number of folds to used

References

Hastie, T., Tibshiani, R., and Friedman, J. (2017), The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition, New York: Springer.

James, G., Witten, D., Hastie, T., and Tibshirani, R. (2013), An Introduction to Statistical Learning with Applications in R, New York: Springer.

Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.

Author(s)

John Kloke

See Also

knn

Examples

train_set <- sim_class2[sim_class2$train==1,-1]
set.seed(19180511)
fit_cv <- knn_cv(train_set,k.cv=10)
fit_cv