An Improved k-Nearest Neighbor Classification Using Genetic Algorithm

k-Nearest Neighbor (KNN) is one of the most popular algorithms for pattern recognition. Many researchers have found that the KNN algorithm accomplishes very good performance in their experiments on different data sets. The traditional KNN text classification algorithm has three limitations: (i) calculation complexity due to the usage of all the training samples for classification, (ii) the performance is solely dependent on the training set, and (iii) there is no weight difference between samples. To overcome these limitations, an improved version of KNN is proposed in this paper. Genetic Algorithm (GA) is combined with KNN to improve its classification performance. Instead of considering all the training samples and taking k-neighbors, the GA is employed to take k-neighbors straightaway and then calculate the distance to classify the test samples. Before classification, initially the reduced feature set is received from a novel method based on Rough set theory hybrid with Bee Colony Optimization (BCO) as we have discussed in our earlier work. The performance is compared with the traditional KNN, CART and SVM classifiers.