The task of classification in applications of data mining is also known as supervised learning where some specific classes are predefined for the training sample and objects are assigned with appropriate class. This is modeled with any classifier and tested on testing data to find the appropriate class. A simple and efficient algorithm for supervised classification is k-Nearest Neighbor (KNN) in which determining the optimal value of k has been an interesting research problem. In this paper, we develop a new algorithm called “Integrated Parallel k-Nearest Neighbor” using the ensembling technique and obtained better results than any other single classifier based on a suitable distance measure and based on any value of k ranging from 1 to 20.
CITATION STYLE
Agrawal, R. (2019). Integrated parallel K-nearest neighbor algorithm. In Smart Innovation, Systems and Technologies (Vol. 104, pp. 479–486). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-13-1921-1_47
Mendeley helps you to discover research relevant for your work.