Image of PENERAPAN AUGMENTASI DAN ENSEMBLE LEARNING PADA ARSITEKTUR INCEPTIONV3, MOBILENET, DAN VISUAL GEOMETRY GROUP 19 UNTUK KLASIFIKASI GANGGUAN DIABETIC RETINOPATY

Skripsi

PENERAPAN AUGMENTASI DAN ENSEMBLE LEARNING PADA ARSITEKTUR INCEPTIONV3, MOBILENET, DAN VISUAL GEOMETRY GROUP 19 UNTUK KLASIFIKASI GANGGUAN DIABETIC RETINOPATY

Penilaian

0,0

dari 5
Penilaian anda saat ini :  

Deep learning has developed rapidly and provides good performance in the image classification process in diabetic retinopathy (DR) disorders. The dataset that is often used in classification is STARE. Unfortunately, only 20 images have labels, therefore augmentation was chosen to overcome data limitations. The high-performance classification method is CNN, so the architecture used is InceptionV3, MobileNet, and VGG19. Unfortunately, in a single classification, overfitting problems often occur during the training process. In this study, the Ensemble learning method will be used to optimize the performance results of the architecture. This study aims to determine the results of the evaluation of the methods used in classifying DR. The stages carried out are data collection, pre-processing, augmentation using rotation and image flipping methods, training, testing, and evaluation of each architecture. The results of the study obtained 18000 new data from augmentation. Ensemble learning performance results outperformed single classification performance with values of accuracy, specificity, sensitivity, F1-Score, and Cohen's kappa which were obtained respectively 95.5%, 95.7%, 95.4%, 95.5%, and 0.911. From these results, it can be said that the Ensemble learning method can perform DR classification very well, which is indicated

Availability
Inventory Code Barcode Call Number Location Status
2107005444T68916T689162021Central Library (Referens)Available
Detail Information
Series Title
-
Call Number
T689162021
Publisher
Indralaya : Jurusan Matematika, Fakultas Matematika Dan Ilmu Pengetahuan Alam, Universitas Sriwijaya.,
Collation
xi, 94 hlm.; Ilus.; 29 cm
Language
Indonesia
ISBN/ISSN
-
Classification
519.507
Content Type
-
Media Type
-
Carrier Type
-
Edition
-
Subject(s)
Jurusan Matematika
Matematika Statistikal
Specific Detail Info
-
Statement of Responsibility
Other version/related

No other version available

File Attachment
Comments
You must be logged in to post a comment