This study aims to explore and develop a K-Nearest Neighbors (KNN)-based classification model using various distance calculation methods, namely Euclidean, Manhattan, Minkowski, and Hamming Distance. To improve the model’s accuracy, the results from each distance method are combined using a weighted average technique. The datasets used are the Iris and Breast Cancer datasets obtained from the UCI Machine Learning Repository. Preprocessing is carried out using normalization with StandardScaler to ensure uniform feature scaling. The model is tested using cross-validation techniques and evaluated using accuracy metrics and a confusion matrix to assess classification performance. Based on the experimental results, the use of multiple distance methods combined with a weighted average approach yields improved accuracy compared to the conventional KNN method that relies on a single distance calculation. The findings of this study indicate that the combination of distance methods in KNN can enhance model performance in classification tasks. This study is expected to contribute to the development of a more adaptive KNN algorithm tailored to diverse data characteristics.