Skripsi

SOFT VOTING ENSEMBLE LEARNING YOLO DENGAN FINETUNING CNN RESNET, EFFICIENTNET DAN DENSENET DALAM DETEKSI DAN KLASIFIKASI CITRA

---

The rapid advancement of computer vision and deep learning has greatly improved image classification performance. However, maintaining consistent accuracy across diverse datasets and visual conditions remains challenging. This study proposes a soft voting ensemble learning approach for image classification using three fine-tuned Convolutional Neural Networks (CNNs): ResNet, EfficientNet, and DenseNet. Each model extracts deep features and produces class probability predictions, which are combined through soft voting to enhance accuracy and robustness. Data augmentation techniques such as rotation, flipping, cropping, brightness adjustment, and resizing were applied to improve generalization. Experimental results show that individual CNNs achieved high accuracy, but some misclassifications persisted. The ensemble method effectively reduced these errors and achieved superior overall accuracy compared to single-model approaches. These findings demonstrate that ensemble learning enhances reliability and generalization in image classification. Future work will evaluate the model on larger and more diverse datasets under real-world lighting and background conditions.

---

Availability

Detail Information

Series Title

-

Call Number

T1945622025

Publisher

Indralaya : Prodi Magister Ilmu Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2026

Collation

92 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

006.307

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Kecerdasan Buatan
Prodi Magister Ilmu Komputer

Specific Detail Info

-

Statement of Responsibility

TUTI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment