CD Skripsi

Implementasi Deep Learning Untuk Identifikasi Daun Tanaman Obat Menggunakan Metode Transfer Learning

---

ABSTRACT
Medicinal plants are a type of plant that have properties for use as an alternative medicine for curing or preventing various diseases. The utilization of medicinal plants in Indonesia has been widely used by the community since ancient times. Knowledge about medicinal plants has also been passed down by ancestors since ancient times. Medicinal plants have leaf shapes that are almost similar between one plant and another, especially in terms of leaf morphology. This makes some people confused in identifying medicinal plant leaves. In recent decades, deep learning has been used to identify images. Deep Learning has the ability to accurately identify objects and is very suitable for identifying medicinal plant leaves. In this research, transfer learning method is used to identify medicinal plants, where transfer learning uses a model that has been previously trained and used as a reference for new tasks. The pretrained model used in this study is MobileNetV2. Fine-tuning techniques were applied in this study to improve model performance. Several experiments were conducted with different parameters such as epoch and fine-tune layers to obtain the best results. The results of this study obtained 99% training accuracy, 98% validation accuracy, and 94% testing accuracy.
Keyword : Medicinal Plants, Deep Learning, Transfer Learning, MobileNetV2

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

1907156542

Penerbit

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

Deskripsi Fisik

-

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

1907156542

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