Enhanced Classification of Iris Species using KMedoids with Explainable AI

In this paper, we propose a K-mediod clustering algorithm for classification of iris data into different species like Setosa, Versicolor, and Virginica, respectively. The proposed algorithm is used to clustering by finding the mediods. Moreover, XAI methods like PDP and SHAP interaction values are used to present the decision-making and feature importance. The analyses reveal increased precision and sensitivity of searches with a desirable level of specificity preserved. The proposed model is compared with that of K-Nearest Neighbor and Threshold clustering, respectively. The parameters such as accuracy, precision, sensitivity and specificity are considered to observe the performance of the algorithms. The proposed method obtained the accuracy of 90%, sensitivity of 90%, and precision 92%, respectively.