Article Per Year (5 Year)
p-Index From 2021 - 2026
0
P-Index
This Author published in this journals
All Journal International Conference on Industrial Revolution for Polytechnic Education
Wahyudi Setiawan ,Moh. Imam Utoyo ,Riries Rulaningtyas
Unknown Affiliation
Agriculture, Biological Sciences & Forestry Automotive Engineering Chemical Engineering, Chemistry & Bioengineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Electrical & Electronics Engineering Industrial & Manufacturing Engineering Mathematics

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Transfer Learning using Modification Convolutional Neural Network Model for Classification of Neovascularization Wahyudi Setiawan ,Moh. Imam Utoyo ,Riries Rulaningtyas
International Conference on Industrial Revolution for Polytechnic Education Vol. 2 No. 2 (2020): International Conference on Industrial Revolution for Polytechnic Education
Publisher : PolinemaPress

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Transfer learning is a technique for reusing Neural Network models. Transfer learning makes the system have a knowledge without doing learning first. Transfer Learning can be applied to the Convolutional Neural Network (CNN) models. In this study, the classification of retinal images was carried out into two classes namely Normal and Neovascularization. The test was conducted using two scenarios. First, testing uses transfer learning with CNN models. CNN models used are AlexNet, VGG16, VGG19, GoogleNet and ResNet50. Second, the experiment uses the CNN AlexNet model and classification method. The classification methods used are Support Vector Machine, k-Nearest Neighbors, Naïve Bayes, LogitBoost and Discriminant Analysis. The experiment data uses public data from the MESSIDOR and Retina Image Bank. The results of the first scenario showed the highest accuracy of 100%. The second scenario test results have the highest accuracy of 96.43%.
Co-Authors