In _______, each new data item is compared with existing ones using a distance metric to search an existing item that is the closest to the new item.
1) decision tree
2) clustering
3) linear regression
4) instance-based learning