Reachable Distance Function for KNN Classification

Distance function is a main metrics of measuring the affinity between two data points in machine learning. Extant distance functions often provide unreachable values real applications. This can lead to incorrect measure points. paper proposes reachable for KNN classification. The not geometric direct-line It gives consideration class attribute training dataset when Concretely speaking, includes their center and distance. Its shape looks like “Z,” we also call it Z function. In this way, same always stronger than that different classes. Or, intraclass are closer those interclass We evaluated with experiments, demonstrated proposed achieved better performance