Question

True/False With a KNN classifier, training is relatively fast and prediction is relatively slow. In the figure below, the ** symbol represents a point to be classified, the 'x' and 'o' symbols represent 2-D feature vectors of class o and x, and the concentric circles around the central *** point are there to help judge distances from the *** point. In KNN classification, with k=3, what class should be assigned to the central *** point? (write 'x' or '0') o o In KNN anomaly detection, if the threshold value is fixed, then increasing k will tend to give a. more anomalies b. less anomalies no change in the number of anomalies (write a, b, or c) C.

Answer 1

For k=3 , out of 3 nearest neighbors, there are two 'o'.

Therefore the central point will be classified as 'o.

In KNN anamoly detection if the threshold value is fixed then increasing k will tend to give --> less anamolies

comment for detailed explanation ow do upvote.