k‐Nearest Neighbor Algorithm

The most common data mining task is that of classification tasks that may be found in nearly every field of endeavor: banking, education, medicine, law and homeland security. The author investigates k‐nearest neighbor algorithm, which is most often used for classification task, although it can also be used for estimation and prediction. Data analysts define distance metrics to measure similarity of records. Combination function combines these similar records to provide a classification decision for the new record. For instance‐based learning methods such as the k‐nearest neighbor algorithm, it is vitally important to have access to a rich database full of as many different combinations of attribute values as possible. The chapter discusses application of the k‐nearest neighbor algorithm using IBM/SPSS modeler, and defines the term stretching the axes.