Application of K-Nearest Neighbor Algorithm for Estimating Fishery Product Quality at BPPMHKP Pontianak

Microbiological quality is a key parameter in ensuring the safety and overall integrity of fishery products. One of the primary indicators in microbiological testing is the Total Plate Count (TPC), locally known in Indonesia as Angka Lempeng Total (ALT), which measures the concentration of aerobic microorganisms in a sample. This study aims to develop a predictive system for ALT values in fishery products using the K-Nearest Neighbor (KNN) algorithm as the main classification approach. The dataset was derived from microbiological test results conducted at the BPPMHKP Laboratory in Pontianak between 2020 and 2024. Data preprocessing included converting ALT values from scientific text format to numeric values, applying Min-Max normalization, splitting the dataset into training and testing subsets, and implementing the KNN algorithm with K = 3. Predictions were generated by calculating the Euclidean distance between each test sample and the training set, selecting the three nearest neighbors, and averaging their ALT values. The proposed system achieved a prediction accuracy of 98.66% compared to actual ALT measurements. Based on the microbiological threshold of 5.0 × 10ⵠcolony-forming units per gram (CFU/g) used by BPPMHKP, the system effectively estimated product quality according to safety standards. These findings indicate the potential of the system to be developed into an application-based decision-support tool for government laboratories and quality control agencies. Keywords: Total Plate Count, K-Nearest Neighbor, Prediction, Fishery Products, Accuracy