CAS Exam MAS-II Fall 2018, Question 37
Study guide: Exam MAS-II
You are given a data set consisting of 150 data points. There are three possible classifications for each data point, as shown in the leftmost graph, with 50 data points falling into each classification.
Based on this data you train a K-Nearest Neighbors model using 95 of the data points. You evaluate k ranging from 1 to 95. For each k, you calculate the test error rate on the remaining 55 data points with results shown in the rightmost graph.
Determine the cause of the rapid increase in error rate between k equals 70 and k equals 95.
- A) As k increases the K-Nearest Neighbors algorithm performs better.
- B) As k increases the K-Nearest Neighbors algorithm performs worse.
- C) As k approaches 95, all data points are predicted to have the same classification.
- D) As k approaches 95, all data points are incorrectly classified.
- E) There is no clear relationship between the value of k and the error rate.
Answer this question and see the worked solution