A pairing individual-trades system, using KNN method: The educational and vocational guidance as a case study
Essaid El Haji, Abdellah Azmani, Mohamed El Harzli · 2014
This paper presents a pairing system, individualtrades, based on the supervised classification method k-nearest neighbors (KNN). This method consists in determining, for each new observation to be classified, the list of nearest neighbors of the observations already classified. The observation is assigned to the class that contains the largest number of observations among the nearest neighbors. The use of the KNN method requires choosing a distance and the most classical one is the Euclidean distance. In the context of this work, we will test two functions to measure resemblance as far as similarity and dissimilarity are concerned.