Classification of Potato Species using Transfer Learning and Support Vector Machine

Abstract

In the last decade, the importance of artificial intelligence applications has increased. The potato has become one of the most consumed vegetables in the world. In this study, it has aimed to classify potato species by using Transfer Learning with Visual Geometry Group-19 (VGG-19), which is a kind of Convolutional Neural Network (CNN) and support vector machine (SVM) which is a powerful Machine Learning algorithm has been used. This private dataset consists of 400 images, whose original pixels are 2752x5664, with 100 images per class. Utilized the power of transfer learning on VGG-19, which has %73,3 overall accuracy, and support vector machine, which has %71,6 accuracy, which means transfer learning with VGG-19 is better than SVM by %1,7. This work demonstrates the performance of Machine Learning models in a private dataset.