Sultoni, Royan Fajar (2024) Analisa Komparasi Algoritma MachineLearning dan Deep Learning dalam Klasifikasi Citra Ras Kucing. Undergraduate thesis, UPN Veteran Jawa Timur.
Text (Cover)
20081010175-cover.pdf Download (813kB) |
|
Text (BAB 1)
20081010175-bab1.pdf Download (62kB) |
|
Text (Bab 2)
20081010175-bab2.pdf Restricted to Repository staff only until 19 July 2026. Download (399kB) |
|
Text (Bab 3)
20081010175-bab3.pdf Restricted to Repository staff only until 19 July 2026. Download (434kB) |
|
Text (Bab 4)
20081010175-bab4.pdf Restricted to Repository staff only until 19 July 2026. Download (1MB) |
|
Text (Bab 5)
20081010175-bab5.pdf Restricted to Repository staff only until 19 July 2026. Download (51kB) |
|
Text (Daftar Pustaka)
20081010175-daftarpustaka.pdf Download (123kB) |
|
Text (Lampiran)
20081010175-lampiran.pdf Restricted to Repository staff only Download (172kB) |
Abstract
Cat (Felis catus) is a type of carnivorous mammal from the family Felidae that has been domesticated and has lived alongside humans since ancient times. Domestic cats are broadly classified into two types: stray cats and pedigree cats. Pedigree cats have a wide variety of breeds, which often leads to confusion in identifying the breed of a cat. In practice, each breed requires different treatment, especially in terms of care. In digital image processing, Machine Learning and Deep Learning are key aspects in applying technology to address this issue, leading to the design of research related to this problem. This research aims to contribute to further studies in more advanced and effective image recognition processes. In the experiments conducted in this study, SVM, KNN, and CNN methods with Xception and EfficientNet-B1 architectures were tested. Based on the final results obtained from this testing, the CNN method with the Xception architecture proved to be the best model. By using fine-tuning and a learning rate of 1e-5, this method achieved a micro average value of 0.974 on a dataset of 13 cat breeds and 7800 images. Meanwhile, the method that achieved the fastest ETA (Estimated Time of Arrival) for Training and Testing was the KNN method, with an ETA Training time of 0.194 seconds and an ETA Testing time of 1.782 seconds.
Item Type: | Thesis (Undergraduate) | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Contributors: |
|
||||||||||||
Subjects: | T Technology > T Technology (General) T Technology > T Technology (General) > T385 Computer Graphics |
||||||||||||
Divisions: | Faculty of Computer Science > Departemen of Informatics | ||||||||||||
Depositing User: | Royan Fajar Sultoni | ||||||||||||
Date Deposited: | 19 Jul 2024 08:59 | ||||||||||||
Last Modified: | 19 Jul 2024 08:59 | ||||||||||||
URI: | https://repository.upnjatim.ac.id/id/eprint/26760 |
Actions (login required)
View Item |