Classification of Flower Dataset using Machine Learning Models

Abstract

Modern day machine learning aims to categorize data based on developed models and predict future outcomes according to these models. Today Machine Learning finds its application in various fields such as facial recognition, speech recognition, medical diagnosis for example predicting potential heart failure, sentiment analysis, product recommendations etc. This paper proposes 3 classification models to efficiently predict the Iris flower species. The proposed model uses Exploratory Data Analysis (EDA) to analyse and pre-process the dataset and the prediction is performed by the three classification models namely- "Logistic Regression, Support Vector Machine (SVM) and K-Nearest Neighbours (KNN)". All the proposed models are tested on Iris dataset and achieved maximum accuracy of 96.43, 98.21 and 94.64 percent respectively. This paper provides a thorough analysis of the various supervised machine learning models that are suitable for predicting the species of Iris flower based on the various attributes like sepal width, sepal length, petal width and petal length.