CLASSIFYING IRIS FLOWER DATA USING ALGORITHMS NAÏVE BAYES, RANDOM FOREST AND KNN

Abstract

Iris is a beautiful flower, representing luck and love courage, loyalty, and wisdom. Therefore, the classification and accurate prediction of Iris flower brings many important meanings in practice. Although there have been many scientific publications related to classification and prediction of Iris flowers, the classification and prediction performance of these publications still have certain limitations that need to be studied for further improvement. In this paper, the author proposes model to classify and predict Iris flowers on the basis of the application of the Weka toolkit and the Naïve Bayes, Random Forest and KNN algorithms. The results reveal that all three algorithms above give high accuracy (over 95%), so it is suitable for building model to classify Iris flowers. However, the two algorithms, Random Forest and KNN (k=3), show better stability and objectivity than the Naïve Bayes algorithm.