Comparison of k-NN and Naive Bayes Algorithms for Classifying Mackerel Tuna Freshness For Real-Time Classification Using Gas Sensors

Setyagraha , Muhammad Rafi Mahfuz
Novamizanti, Ledya
Wijaya, Dedy Rahman

eProceedings of Engineering
Vol. 12 No. 2 (2025): April 2025

The large production and consumption of mackerel tuna in Indonesia reflect its importance as a local staple and a valuable export product contributing to the nation's economy. Mackerel tuna is prized for its nutritional content and affordability, making it a crucial part of the diet for many Indonesians. Ensuring the freshness and quality of this high-demand product is essential. This study introduces a machine-learning approach to detect fish freshness by analyzing gases emitted during spoilage, utilizing MQ-2, MQ-9, and MQ-135 gas sensors. The data were processed using the k-Nearest Neighbors (k-NN) and Naive Bayes algorithms, both achieving accuracy rates near 100%. These findings highlight the system’s potential to enhance quality control in Indonesia’s fishery industry by offering an efficient and reliable method for assessing fish freshness. Keywords—classification, machine learning, tuna, gas sensor