Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Refining Parameter Values in Convolutional Neural Networks Using Weight Modification Particle Swarm Optimization for Enhanced Coral Reef Condition Classification

Rektorat UDINUS 2024 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2024
Goals
Record
177ce5e5-00f8-46b7-89c4-cb350283ff8f
Published
2026-10-30

Universitas Dian Nuswantoro (UDINUS), through the participation of Andono, Pulung Nurtantio; Rohman, Muhammad Syaifur; Megantara, Rama Aria; Pergiwati, Dewi; Wahyudi, Firman; Shidik, Guruh Fajar; and Pramuendar, Ricardus Anggi, conducted a study on refining parameter values in Convolutional Neural Networks (CNN) for coral reef condition classification. The activity addressed the context that Indonesia has a vast marine area with numerous coral reefs located at the center of the global coral triangle and that these reefs are crucial for aquatic life. The source identifies climate change, pollution, rising sea temperatures, and human activities as severe threats to coral reefs, with 85% of Indonesia’s coral reefs stated to be at risk.

The study used Particle Swarm Optimization (PSO) influenced by declining weighting to improve CNN performance by determining parameter values for coral reef condition classification. The optimized parameters included Epoch, Dropout, Hidden Size, Learning Rate, Activation Function, and Optimizer. The best values produced by the proposed method were Epoch 9, Dropout 0.2, Hidden Size 256, Learning Rate 0.002, ReLU as the Activation Function, and Adam as the Optimizer. The experimental results showed more consistent and stable performance than the Original PSO and Original GWO methods. The proposed method achieved a maximum training accuracy of 82.23% and testing accuracy of 79.74%. The source states that the method can be considered more reliable and effective for parameter optimization in CNN-based coral reef classification. This activity relates to climate action and marine ecosystem research by applying computational methods to the monitoring and classification of coral reef conditions.

Reference https://www.scopus.com/pages/publications/85207878823?origin=resultslist