COMPARING THE ACTIVATION OF RELU, ELU, AND TANH FUNCTIONS IN CLASSIFYING BETTA FISH SPECIES USING CNN

RUMONDOR, YOHANES BOSKO KEVIN (2021) COMPARING THE ACTIVATION OF RELU, ELU, AND TANH FUNCTIONS IN CLASSIFYING BETTA FISH SPECIES USING CNN. Other thesis, Universitas Katholik Soegijapranata Semarang.

[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-COVER_a.pdf

Download (1MB) | Preview
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-BAB I_a.pdf

Download (114kB) | Preview
[img] Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-BAB II_a.pdf
Restricted to Registered users only

Download (235kB)
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-BAB III_a.pdf

Download (140kB) | Preview
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-BAB IV_a.pdf

Download (137kB) | Preview
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-BAB V_a.pdf

Download (357kB) | Preview
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-BAB VI_a.pdf

Download (111kB) | Preview
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-DAPUS_a.pdf

Download (231kB) | Preview
[img]
Preview
Text
16.K1.0022-YOHANES BOSKO KEVIN RUMONDOR-LAMP_a.pdf

Download (457kB) | Preview

Abstract

Betta fish is one of the most popular freshwater ornamental fish, especially in Indonesia because in addition to easy maintenance, breeding is also relatively easy. The most common betta fish, especially in Indonesia, are Crowntail, Halfmoon, and Plakat. But some people can't tell the difference between the types. Therefore, in this final project, we will classify the types of betta fish. The Crowntail has fins that resemble spines or fingers, the Halfmoon has very wide fins and the tail is like a half moon, while the plakat is almost the same as the halfmoon, only the fins are small. Therefore, the author uses CNN to classify the types of betta fish. In addition to classifying the types of betta fish, the author will also compare which 3 activations (Relu, Tanh, Elu) have the best results. Based on the results of tests that have been carried out by CNN, it can work well to classify the types of betta fish. By using Relu activation, the test results get an accuracy of 80%.

Item Type: Thesis (Other)
Subjects: 000 Computer Science, Information and General Works > 004 Data processing & computer science
Divisions: Faculty of Computer Science > Department of Informatics Engineering
Depositing User: mr AM. Pudja Adjie Sudoso
Date Deposited: 14 Oct 2021 06:01
Last Modified: 14 Oct 2021 06:01
URI: http://repository.unika.ac.id/id/eprint/27116

Actions (login required)

View Item View Item