Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2024
- Goals
- Record
- 2575ae7a-1e69-480d-90af-0e0f2ef77618
- Published
- 2026-10-30
Universitas Dian Nuswantoro conducted research on optimizing early coronary heart disease prediction through a genetic algorithm-based convolutional neural network feature engineering approach. The activity involved Hidayat, Erwin Yudi; Astuti, Yani Parti; Dewi, Ika Novita; Salam, Abu; Soeleman, Moch. Arief; Hasibuan, Zainal Arifin; and Yousif, Ahmed Sabeeh as participants. The research addressed limitations of traditional hyperparameter optimization techniques in coronary heart disease detection and examined whether a genetic algorithm could support improved predictive performance.
The method used a genetic algorithm for hyperparameter optimization. This process navigated a complex combinatorial space to identify configurations for a convolutional neural network model. Information gain was employed for feature selection optimization, and coronary heart disease datasets were transformed into image-like input for the convolutional neural network architecture. The efficacy of the approach was benchmarked against traditional optimization strategies. The source describes the work as focused on binary and multiclass coronary heart disease prediction tasks.
The findings reported that the advanced genetic algorithm-based convolutional neural network model outperformed traditional methods and achieved an increase in accuracy. The optimized model delivered a peak accuracy of 85% in hyperparameter optimization. When integrated with naïve Bayes, support vector machine, decision tree, logistic regression, and random forest, the model achieved 100% accuracy for both binary and multiclass coronary heart disease prediction tasks.
The research concluded that integrating a genetic algorithm into convolutional neural network feature engineering improved coronary heart disease prediction accuracy. The approach was described as having a high degree of predictive reliability and as potentially contributing to artificial intelligence-driven healthcare, with the possibility of clinical deployment for early coronary heart disease detection. Future work will focus on expanding the approach to a wider set of coronary heart disease data and considering potential integration with wearable technology for continuous health monitoring. The activity provides institutional evidence of research addressing health-related prediction through artificial intelligence methods.