In any nearesr nighbour problem :
- We take a value from test set
- apply distance formula on this point in test set to every point in train set
3)Take the minimum distance value from the test point and take this points index
- get the test point at which the distance is low with help of this index
- Get the label of the test point using train_label(which is the label data of training data)
6)compare this label with test_label(which is part of test data)
- If there are equal it correct prediction and error rate lowers,if not error increases.
Are these steps correct,if not explain me please.