Skip to content
ml-recall

#knn

15 questions from Evaluation Metrics, Nearest Neighbors & Kernel Methods, Clustering

Timed test

15 questions · 15 min · score and explanations at the end

Take a test
Q 1 / 15Score 0 / 0

Question 1: Short answer

mediumAI-checkedEvaluation Metrics

A dataset of 1000 cases is split into 600 training and 400 validation cases. A 1-nearest-neighbour classifier has an 8% error rate on the validation set. You then discover that 100 training cases had accidentally been copied over 100 validation cases, so 100 validation cases are exact duplicates of training cases. What is the error rate on the 300 genuine validation cases?

a fraction between 0 and 1, 3 decimals