Text

PENERAPAN METODE ENSEMBLE LEARNING PADA HASIL U-NET DAN RESNET MENGGUNAKAN TEKNIK WEIGHTED VOTING UNTUK SEGMENTASI AKAR PADA TANAH

---

Penilaian

---

The feature learning capability using Convolutional Neural Network (CNN) has good performance in image segmentation. The architecture of CNN which is often used for image segmentation is U-Net and ResNet architecture. Several studies have proposed the Ensemble Learning (EL) method to combine the results of architectural performance in each image segmentation. Weighted Voting is one of the most frequently used EL methods, Weighted Voting works by choosing the largest weight from the final prediction generated by each segmentation model. The dataset used is the root image of Cichorium intybus L (Chicory) taken from the Zenodo dataset with performance evaluation measures namely accuracy, precision, sensitivity, specificity, F1-Score and Intersection over Union (IoU). The results of this study obtained an accuracy of 99.93%, precision 94.86%, sensitivity 85.19%, specificity 99.98%, F1-Score 89.77%, and IoU 81.44%. Based on these result, it shows that the proposed architecture is able to perform root segmentation in the soil well based on the category of model performance evaluation.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2207004385	T78319	T783192022	Central Library (Referens)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T783192022

Publisher

Inderalaya : Jurusan Matematika, Fakultas Matematika Dan Ilmu Pengetahuan Alam, Universitas Sriwijaya., 2022

Collation

xii, 56 hlm.; ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

510.285 07

Content Type

-

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Jurusan Matematika
Pengolahan dan Analisa Data di Bidang Matematika

Specific Detail Info

-

Statement of Responsibility

SEPTA

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment