Klasifikasi Flora Endemik Kalimantan Di Kebun Raya Balikpapan Menggunakan EfficientNetV2 - Submit Jurnal

Putro, Dhio Anugrah Prakasa (2026) Klasifikasi Flora Endemik Kalimantan Di Kebun Raya Balikpapan Menggunakan EfficientNetV2 - Submit Jurnal. Bachelor thesis, Institut Teknologi Kalimantan.

[img] Text
11221004_cover.pdf
Download (373kB)
[img] Text
11221004_statement_of_authenticity.pdf
Download (126kB)
[img] Text
11221004_publishing_agreement.pdf
Download (150kB)
[img] Text
11221004_approval_sheet.pdf
Download (4MB)
[img] Text
11221004_preface.pdf
Download (234kB)
[img] Text
11221004_abstract_id.pdf
Download (144kB)
[img] Text
11221004_abstract_en.pdf
Restricted to Repository staff only until 4 October 2028.
Download (136kB) | Request a copy
[img] Text
11221004_table_of_content.pdf
Restricted to Repository staff only until 4 October 2028.
Download (217kB) | Request a copy
[img] Text
11221004_illustrations.pdf
Restricted to Repository staff only until 4 October 2028.
Download (125kB) | Request a copy
[img] Text
11221004_tables.pdf
Restricted to Repository staff only until 4 October 2028.
Download (121kB) | Request a copy
[img] Text
11221004_notations.pdf
Restricted to Repository staff only until 4 October 2028.
Download (128kB) | Request a copy
[img] Text
11221004_chapter_1.pdf
Restricted to Repository staff only until 4 October 2028.
Download (626kB) | Request a copy
[img] Text
11221004_chapter_2.pdf
Restricted to Repository staff only until 4 October 2028.
Download (2MB) | Request a copy
[img] Text
11221004_chapter_3.pdf
Restricted to Repository staff only until 4 October 2028.
Download (2MB) | Request a copy
[img] Text
11221004_chapter_4.pdf
Restricted to Repository staff only until 4 October 2028.
Download (1MB) | Request a copy
[img] Text
11221004_conclusions.pdf
Restricted to Repository staff only until 4 October 2028.
Download (234kB) | Request a copy
[img] Text
11221004_bibliography.pdf
Download (332kB)
[img] Text
11221004_enclosure.pdf
Restricted to Repository staff only until 4 October 2028.
Download (138kB) | Request a copy
[img] Text
11221004_paper.pdf
Restricted to Repository staff only until 4 October 2028.
Download (451kB) | Request a copy
[img] Text
11221004_presentation.pdf
Restricted to Repository staff only until 4 October 2028.
Download (1MB) | Request a copy
[img] Text
11221004_Form. TA-020.pdf
Restricted to Repository staff only until 4 October 2028.
Download (39kB) | Request a copy

Abstract

Kalimantan sebagai pusat megabiodiversitas memiliki kekayaan flora endemik yang penting bagi ekosistem, namun proses identifikasinya masih dilakukan secara manual dan bergantung pada pakar. Penelitian ini bertujuan mengembangkan model klasifikasi otomatis flora endemik Kalimantan menggunakan metode transfer learning dengan arsitektur EfficientNetV2 berbasis dataset citra dari Kebun Raya Balikpapan. Hasil penelitian menunjukkan bahwa seluruh model mampu mencapai performa tinggi dengan akurasi dan F1-Score di atas 95%, di mana EfficientNetV2-L menjadi model terbaik dengan akurasi sebesar 97,58% dan F1-Score sebesar 97,59% serta waktu inferensi sebesar 0,1537 detik, sedangkan EfficientNetV2-S memiliki waktu inferensi tercepat sebesar 0,099 detik dengan performa yang lebih rendah. Berdasarkan analisis Pareto optimal, EfficientNetV2-L menjadi model yang paling direkomendasikan karena memberikan peningkatan akurasi yang signifikan dengan tambahan waktu inferensi yang relatif kecil, sementara EfficientNetV2-M tidak direkomendasikan karena berada di luar garis Pareto dan terdominasi. Dengan demikian, arsitektur EfficientNetV2 terbukti efektif untuk klasifikasi flora endemik Kalimantan dan berpotensi mendukung proses dokumentasi serta konservasi berbasis kecerdasan buatan.

Item Type: Thesis (Bachelor)
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QK Botany
Divisions: Jurusan Matematika dan Teknologi Informasi > Informatika
Depositing User: Dhio Anugrah Prakasa Putro
Date Deposited: 16 Jul 2026 06:27
Last Modified: 16 Jul 2026 06:27
URI: http://repository.itk.ac.id/id/eprint/27151

Actions (login required)

View Item View Item