Analysis of Different Classification Algorithms Used for the Classification on Three Species of Iris Limniris (Tausch) Spach Dataset

Iris is a flowering plant having 5-6 sepals which is a characteristic feature of classification of plant species. To determine the species of this genus the number of sepals are one of the factors and this work provides in an easier way to classify the Iris species. Hence, this paper explores the analysis of commonly used machine learning supervised classification algorithms on classifying and predicting three Iris flowering plant species i.e., Iris setosa, Iris virginica, Iris versicolor from the iris flower dataset present in UCI Machine Learning Repository which was compiled by Ronald Fisher. The dataset contains data of sepal length, sepal width, petal length, and petal width which is used for predicting the required species. Classification algorithms are a subset of supervised learning. Support Vector Machines, Decision Tree Classifier, and Logisitic Regression are the algorithms used in this paper for the purpose. The dataset is analyzed and preprocessed before fitting the algorithms for the prediction using scikit learn library and the data is analyzed using Python language. Machine Learning libraries for python, which include, pandas, numpy, matplotlib, and seaborn are used. The environment used for this project is Google Collaboratory. The parameters are adjusted for the dataset requirements. Finally, the three algorithms are evaluated based on their accuracy score and confusion matrix, where the SVM showed higher accuracy compared to that of the other two algorithms under these parameters. The significance of the prediction helps in predicting the species as well as eliminating the source of human error in separating different species. This work may contribute to prediction of more species of different genera. This paper comes under the theme: Life Sciences, Biomedical Sciences and Biotechnological aspects.