CD Skripsi

Deteksi COVID-19 Berdasarkan Citra Sintetis Chest X-Ray (CXR) Menggunakan Deep Convolutional Generative Adversarial Networks (DCGAN) Dan Transfer Learning

---

ABSTRACT
The global COVID-19 pandemic has had a significant impact on the health and lives of many people worldwide, with a high number of cases and fatalities. The need for rapid and accurate diagnosis is crucial. The use of radiographic imaging, particularly chest radiography (CXR), has been considered for the diagnosis of suspected COVID-19 patients. However, the limited availability of CXR data poses a challenge in developing accurate detection models. In this study, a larger dataset was generated using Deep Convolutional Generative Adversarial Networks (DGAN). The Expanded Dataset consists of 34.63% original images and 65.37% synthetic images, which were then used to train three pre-trained models: ResNet50, EfficientNetV1, and EfficientNetV2. The results of the study showed that the use of synthetic CXR images generated by DCGAN was able to improve the performance of the models, achieving high detection accuracy. Specifically, the EfficientNetV1 model achieved the highest accuracy of 99.21% with only ten epochs and a training time of 13.23 minutes, compared to a previous accuracy of 98.43%.
Keywords: COVID-19, transfer learning, Deep Convolutional Generative Adversarial Networks (DCGAN), CXR image, detection.

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

1907112680

Penerbit

Pekanbaru : Universitas Riau – Fakultas Teknik – Teknik Elektro., 2023

Deskripsi Fisik

-

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

1907112680

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

-

Subjek

TEKNIK ELEKTRO

Info Detail Spesifik

-

Pernyataan Tanggungjawab

DAUS

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar