Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Optimization of Convolutional Neural Networks Using Particle Swarm Optimization for Malaria Disease Cell Classification

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
d84927e2-3da7-428c-80d9-b7642945d3c2
Published
2026-10-30

Universitas Dian Nuswantoro, through the participation of Anisa, Izza; Purwanto; Guruh Fajar Shidik; and Catur Supriyanto, conducted research on optimizing Convolutional Neural Networks for malaria disease cell classification. The activity addressed malaria as a significant global health challenge and focused on early and accurate detection for effective treatment and control.

The research applied Particle Swarm Optimization to optimize a Convolutional Neural Network for malaria classification using microscopy images. Particle Swarm Optimization was employed to automatically evolve the network architecture, including layer configurations and filter sizes, together with the dropout and learning-rate hyperparameters. The experiments used the National Library of Medicine Malaria dataset sourced from Kaggle. The dataset contained blood smear images collected from 201 patients and was classified into malaria-infected and uninfected groups, providing a foundation for model training and evaluation.

The dataset was divided into training, validation, and testing portions to support model assessment. The experimental results reported that the Particle Swarm Optimization-based Convolutional Neural Network outperformed manually tuned Convolutional Neural Networks, offering superior performance and reliability. The proposed method achieved an accuracy of 99.40% and was presented as a promising solution for malaria detection through automated analysis of medical images. The findings suggested that integrating Particle Swarm Optimization with Convolutional Neural Networks could contribute to the advancement of automated healthcare systems. The source does not provide information about implementation beyond the described experiments. This activity is framed institutionally as UDINUS research concerning health research and automated malaria detection.

Reference https://www.scopus.com/pages/publications/105036815859?origin=resultslist