Please use this identifier to cite or link to this item:
https://repositori.uma.ac.id/handle/123456789/23762
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Muhathir | - |
dc.contributor.advisor | Muliono, Rizki | - |
dc.contributor.author | Tumanggor, Amri Ismail | - |
dc.date.accessioned | 2024-04-19T02:17:27Z | - |
dc.date.available | 2024-04-19T02:17:27Z | - |
dc.date.issued | 2023-09-18 | - |
dc.identifier.uri | https://repositori.uma.ac.id/handle/123456789/23762 | - |
dc.description | 46 Halaman | en_US |
dc.description.abstract | Dengan memanfaatkan deep learning, deteksi dan klasifikasi objek yang hampir sama merupakan hal mendasar dan tantangan untuk memberikan akurasi terbaik dalam membedakan dua larva yaitu ulat Jerman dan ulat hongkong. Ulat jerman dan ulat hongkong merupakan larva dengan morfologi yang sama namun ulat jerman lebih bergizi dibandingkan ulat hongkong. Karena kesamaan morfologi kedua larva tersebut pengetahuan masyarakat terutama pecinta hewan terbatas sehingga sedikit sulit membedakanya. Tujuan penelitian ini adalah menyajikan hyperparameter terbaik dalam klasifikasi Ulat Jerman dan Ulat hongkong dengan model arsitektur Xception. Model dilatih dengan menggunakan dataset berupa citra. Parameter yang digunakan dalam pelatihan yaitu Epoch, Batch Size, dan Optimizer. Hasil pengujian menunjukkan Parameter terbaik dan paling optimal dalam klasifikasi Ulat Jerman dan Ulat Hongkong adalah Parameter dengan kombinasi Epoch 5 Batch Size 8 dengan Optimizer SGD dengan akurasi 100% dan waktu komputasi 27,9 menit. By utilizing deep learning, the detection and classification of nearly identical objects are fundamental aspects and challenges to achieve the best accuracy in distinguishing two types of larvae, namely the Zophobas Morio and the Tenebrio molitor. Zophobas morio and Tenebrio molitor share similar morphologies, but Zophobas morio are more nutritious compared to Tenebrio molitor. Due to the morphological similarities between these larvae, public knowledge, especially among animal enthusiasts, is limited, making it somewhat challenging to differentiate them. The purpose of this research is to present the best hyperparameters for the classification of Zophobas morio and Tenebrio molitor using the Xception architecture model. The model was trained using a dataset consisting of images. The parameters used in training are Epoch, Batch Size, and Optimizer. The test results indicate that the optimal and best parameters for the classification of zophobas morio tenebrio molitor are the combination of Epoch 5, Batch Size 8, with Optimizer SGD, achieving an accuracy of 100% and a computational time of 27.9 minutes. | en_US |
dc.language.iso | id | en_US |
dc.publisher | UNIVERSITAS MEDAN AREA | en_US |
dc.relation.ispartofseries | NPM;178160045 | - |
dc.subject | ulat jerman | en_US |
dc.subject | ulat hongkong | en_US |
dc.subject | xception | en_US |
dc.subject | zophobas morio | en_US |
dc.subject | tenebrio molitor | en_US |
dc.subject | zception | en_US |
dc.title | Hyperparameter Model Arsitektur Xception dalam Mengklasifikasi Ulat Jerman dan Ulat Hongkong | en_US |
dc.title.alternative | Hyperparameters of the Xception Architectural Model in Classifying German Caterpillars and Hong Kong Caterpillars | en_US |
dc.type | Skripsi Sarjana | en_US |
Appears in Collections: | SP - Informatic Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
178160045 - Amri Ismail Tumanggor - Fulltext.pdf | Cover, Abstract, Chapter I, II, III, V, Bibliography | 1.7 MB | Adobe PDF | View/Open |
178160045 - Amri Ismail Tumanggor - Chapter IV.pdf Restricted Access | Chapter IV | 394.47 kB | Adobe PDF | View/Open Request a copy |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.