Classifying Flowers Images by using Different Classifiers in Orange

Abstract

This paper presents the first step towards looking for an advanced solution of image classification using distinct Classifiers in the Orange Data Mining Tool. The objective of the paper is to decide the ability of distinct classifiers for flowers image classification involving a small sample; Deep learning models are used to calculate a feature vector for every image of the Iris flower database. The used classifiers involved logistic regression, Neural Network, AdaBoost, Support Vector Machine, Random Forest and K-NN. The result indicates that the Logistic Regression, Neural Network, AdaBoost classifiers perform best in classifying a small sample of Iris flower images, and SVM and Random Forest classifiers perform less classification accuracy then above classifiers while K-NN performs worst with the lowest classification accuracy.

Komparasi Algoritma Support Vector Machine Dan Naive Bayes Pada Klasifikasi Ras Kucing

“…Accuracy (CA) menggambarkan nilai akurasi dan kinerja dari model pengklasifikasi yang bernilai benar. CA adalah rasio prediksi yang benar terhadap data latih keseluruhan [8].…”

“…Precision adalah rasio prediksi positif yang benar yang dibandingkan dengan hasil positif keseluruhan [8].…”

“…Recall adalah rasio prediksi positif yang benar terhadap keseluruhan data yang merupakan rasio positif yang benar [8].…”