Andi Daniah Pahrany, Edy Saputra Rusdi, Aidawayati Rangkuti, Nur Hilal A. Syahrir, Wahyudi Rusdi
Marine biodiversity in Indonesia includes various types of Marine Biota organisms. Marine biota provides a variety of compounds with antibiotic potential that have not been fully explored. This study uses an Artificial Neural Network (ANN) classification method to identify marine biota compounds with potential antibiotic properties. The dataset used is 3909 drug compounds (535 antibiotic drug compounds and 3374 non-biotic drug compounds) and 1601 antibiotic compounds utilized for training and validation, while data from 73 marine biota compounds in Indonesian waters were tested. Five resampling methods—Original, SMOTE, Borderline SMOTE, SMOTE TOMEK, and SMOTE ENN—were applied to overcome dataset imbalance during preprocessing. The results are obtained using the SMOTE ENN method to the data, resulting in the highest evaluation matrix value for all cases applied. Accuracy 99.065%, Precision 99.042%, Recall 99.042%, and f1-score 99.061%. Of the 73 marine biota compounds tested in the model, 10 were detected as potential antibiotics. © 2025 American Institute of Physics Inc.. All rights reserved.
Department of Mathematics, Faculty of Mathematics and Natural Sciences, Universitas Negeri Malang, Malang, Indonesia; Department of Mathematics, Faculty of Mathematics and Natural Sciences, Universitas Hasanuddin, Makassar, Indonesia; Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Sulawesi Barat, Sulawesi Barat, Indonesia; Department of Sharia Economics, Faculty of Islamic Economics and Business, State Islamic Institute (IAIN) Sultan Amai Gorontalo, Gorontalo, Indonesia
