View article

KNN (K Nearest-neighbor Classification) is a lazy learning classification algorithm, where it only memorizes the training dataset instead of providing a defined discriminative function. KNN tends to search the nearest neighbor(s) for a target in the entire training set, hence, the prediction step of KNN is quite time consuming. KD-tree (K Dimensional-tree) is a multi-dimensional binary tree, which is a specific storage structure for efficiently representing training data. Therefore, the paper takes the advantages of KNN and KD-tree and then proposes a new classification algorithm called KNN-KD-tree. Eleven datasets have been adopted to conduct experiments. The experiments have shown that the proposed KNN-KD-tree algorithm can efficiently reduce time complexity and significantly improve search performance.