https://medium.com/30-days-of-machine-learning/day-3-k-nearest-neighbors-and-bias-variance-tradeoff-75f84d515bdb

University of Michigan - Ann Arbor

The n_neighbors parameter specifies the number of neighbors to consider when classifying points. This number is sometimes referred to as "K". A high n_neighbors risks overfitting the test data and being poor at predicting new data (high bias). On the other hand, a low n_neighbors risks underfitting the data, hence oversimplifying the model and resulting in a "jumpy" plot of predictions (high variance). The best value for K will depend on the dataset, so one value for K may work well for some datasets, but not for others.

"n_neighbors" Parameter for KNN Classifier

K-Nearest Neighbors and Bias–Variance Tradeoff

Mean squared error (MSE) is a measure of prediction accuracy for a particular K. It is calculated as implied by its name: take the average of the squared differences between true label and prediction. Note that MSE should be computed for out-of-sample error when determining the best K, i.e. using test rather than training data.

Learn Before

Related