K Nearest Neighbour implementation using Hadoop 1.2.1 Execution : hadoop jar KNN.jar KNN input.txt train.txt output K Nearest Neighbour Algorithm is used in several machine learning problems for predicting the class labels based on 'k' nearest neighbours of the given data point. Algorithm : a. Calculate the distance between the input data point and the points given in the training set. b. Sort the distance vector and select top 'k' points which are nearest (least distance) to the input point. c. Output the majority among the class labels of 'k' points selected in step b. MapReduce Procedure : a. Distance of the points and 'k' nearest points are selected and the distance along with class labels are sent to the reduce part. b. Further the output is predicted based on the majority in the input of the reduce part.