Please use this identifier to cite or link to this item:
https://repositori.uma.ac.id/handle/123456789/22249
Title: | Analisis Arsitektur Deep Learning VGG untuk Klasifikasi Jenis Jamur |
Other Titles: | VGG Deep Learning Architecture Analysis for Classification of Fungal Types |
Authors: | Syuhada, Rahmad |
metadata.dc.contributor.advisor: | Muhatir |
Keywords: | klasifikasi jamur;deep learning;VGG-19;mushroom classification |
Issue Date: | Sep-2023 |
Publisher: | Universitas Medan Area |
Series/Report no.: | NPM;198160001 |
Abstract: | Indonesia merupakan salah satu pusat keanekaragaman hayati,Salah satu keanekaragaman jenis tumbuhan adalah jamur. Jamur adalah tumbuhan yang sangat sederhana, memiliki inti, spora, tanpa klorofil, berbentuk sel atau filamen bercabang dengan dinding selulosa atau khitin atau keduanya, Identifikasi jamur masih sulit dilakukan karena banyaknya jenis jamur, kurangnya pengetahuan tentang jamur, dan kurangnya ahli dalam bidang jamur. Selain itu, sebagian besar jamur memiliki tingkat kesamaan yang tinggi dalam karakteristik tertentu, yang menyulitkan dalam mengidentifikasi jenis jamur secara visual oleh manusia yang tidak memiliki keahlian khusus. Oleh karena itu, penting untuk dapat mengklasifikasikan jenis jamur sehingga diharapkan masyarakat lebih paham tentang jenis masing-masing jamur. Penelitian ini menggunakan pendekatan transfer learning dengan arsitektur VGG-19 untuk menyediakan metode yang akurat dalam mengklasifikasi jenis jamur. Terdapat 18 skenario model yang di training dan diperoleh performa model terbaik menggunakan hyperparameter jumlah epoch 50, batch size 64, dan optimizer SGD mendapat akurasi 77.3% pada proses training. Setelah di uji menggunakan data testing dan di evaluasi menggunakan confusion matrix dan classification report, diperoleh score accuracy sebesar 70%. Indonesia is one of the centers of biodiversity. One of the diversity of plant species is mushrooms. Fungi are very simple plants, having a nucleus, spores, without chlorophyll, in the form of cells or branching filaments with cellulose or chitin walls or both. Identification of fungi is still difficult because of the many types of fungi, lack of knowledge about fungi, and lack of experts in the field of mushrooms. In addition, most fungi have a high degree of comfort in certain characteristics, which poses a challenge in visually identifying the type of mushroom by an unskilled human. Therefore, it is important to be able to classify the types of mushrooms so that it is hoped that people will understand more about the types of each mushroom. This study uses a transfer learning approach with the VGG-19 architecture to provide an accurate method for classifying fungal species. There were 18 model scenarios that were trained and obtained the best model performance using the hyperparameter number of epochs of 50, batch size of 64, and the SGD optimizer got 77.3% accuracy in the training process. After being tested using data testing and evaluated using the confusion matrix and classification report, a score accuracy of 70% was obtained. |
Description: | 72 Halaman |
URI: | https://repositori.uma.ac.id/handle/123456789/22249 |
Appears in Collections: | SP - Informatic Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
198160001 - Rahmad Syuhada Fulltext.pdf | Cover, Abstract, Chapter I, II, III, V, Bibliography | 1.67 MB | Adobe PDF | View/Open |
198160001 - Rahmad Syuhada Chapter IV.pdf Restricted Access | Chapter IV | 1.51 MB | Adobe PDF | View/Open Request a copy |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.