About DJPS    Editorial Board    Advisory Board    Secretary of Journal    Events & News         

 

DJPS

Diyala Journal For Pure Science

Scientific Refereed Journal Published By College of Science - University of Diyala

 

Admin of Website

djps.1402.383B

Diyala Journal For Pure Sciences DJPS

ISRA Impact Factor:3.715

P- ISSN:2222-8373, E-ISSN:2518-9255

Volume 14, Issue 2 , April . 2018

 

 

A Comparison Between SVM and K-NN for classification of Plant Diseases 

 

Sarah Saadoon Jasim , Ali Adel Mahmood Al-Taei

 

Year: 2018, Volume: 14, Issue: 2

 

Pages: 94-105 , DOI: http://dx.doi.org/10.24237/djps.1402.383B

 

Language: English

 

Download :  

 

Abstract


 

Vegetable crops differ in size, shape, and color and which its suffer from this many leaf batches according to a particular reason. As a result of the plant, pathogens happen for Leaf batches. In agriculture whole fructification, it is essential to learn the origin of plant disease bundles early to be prepared for suitable timing control. In this regard, uses Support Vector Machine (SVM) and K- Nearest Neighbor to classify the plant's symptoms according to their appropriate classifications. These typesare (YS) Yellow Spotted class, (WS) White Spottedclass, (RS) Red Spotted class, and (D) tarnishedclass. Results obtained using SVM algorithm was compared with results obtained by a K-NN algorithm. Specifically, the overall accuracy of SVM model is about 88.17% and 85.61% for the k -NN model (with k = 1).

Keywords: Classification, Support Vector Machine (SVM), k- Nearest Neighbor (k-NN).

 

Copyright © 2017 Diyala Journal For Pure Science (DJPS) . All rights reserved
3:45