Analisis Efektivitas Algoritma Komputasi pada Sistem Pendukung Keputusan

Abstract

Decision Support Systems (DSS) play a crucial role in assisting decision-makers by analyzing large and complex datasets to generate actionable insights. The core performance of a DSS relies heavily on the computational algorithms embedded within its structure, which are responsible for data processing, pattern recognition, and prediction. This study aims to evaluate the effectiveness of three commonly used algorithms Decision Tree (C4.5), Naive Bayes, and K-Nearest Neighbor (K-NN) in supporting decision-making processes using healthcare-related data. The analysis focuses on three performance metrics: classification accuracy, computational speed, and memory usage. A benchmark dataset on heart disease from the UCI Machine Learning Repository was utilized for empirical testing. Results indicate that the Decision Tree algorithm achieved the highest accuracy (92%) and interpretability, making it well-suited for transparent decision-making contexts. Naive Bayes demonstrated the fastest processing time and lowest memory consumption, making it ideal for real-time or resource-constrained systems. Meanwhile, K-NN showed moderate performance but was sensitive to parameter tuning and data volume. These findings suggest that algorithm selection should be aligned with system requirements and resource availability. The study contributes to the development of more efficient and tailored decision support systems by providing empirical evidence of algorithmic strengths and limitations across multiple evaluation dimensions.