Model Klasifikasi Machine Learning Berbasis Multiple Measurement Distance
DOI:
https://doi.org/10.22303/csrid-.17.2.2025.256-269Keywords:
K-Nearest Neighbors, Euclidean Distance, Manhattan Distance, Minkowski Distance, Hamming Distance, Weighted Average, ClassificationAbstract
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.
References
Zhang et al., 2021. Hybrid Distance Metrics in KNN for Improved Performance. Journal of Machine Learning Research.
Li et al., 2022. Weighted KNN with Adaptive Distance Metrics for Medical Diagnosis. ICML.
IEEE Transactions, 2020. Optimized KNN Algorithm for High-Dimensional Data Classification.
Neural Computing and Applications, 2023. Performance Analysis of KNN with Various Distance Measures in Text Classification. AAAI Conference, 2024. An Ensemble Approach to KNN with Multiple Distance Metrics.
Journal of Biomedical Informatics, 2020. Feature Selection in KNN for Medical Diagnosis.
International Journal of Computer Vision, 2021. Dynamic Weighting in KNN Classification.
IEEE CVPR, 2022. Impact of Distance Metrics on KNN Performance in Image Recognition.
Neural Networks Journal, 2023. A Study on KNN Classification in Large Datasets.
ACM SIGKDD, 2024. Advancements in KNN-Based Recommender Systems.
Zhang, L., Wang, Y., & Xu, C. (2021). Hybrid Distance Metrics in KNN for Improved Performance. Journal of Machine Learning Research, 22(1), 1234–1248.
Li, X., Zhou, J., & Liu, Q. (2022). Weighted KNN with Adaptive Distance Metrics for Medical Diagnosis. In Proceedings of the 39th International Conference on Machine Learning (ICML).
Chen, H., & Zhao, R. (2020). Optimized KNN Algorithm for High-Dimensional Data Classification. IEEE Transactions on Neural Networks and Learning Systems, 31(12), 5215–5225.
Kumar, A., & Srinivas, K. (2023). Performance Analysis of KNN with Various Distance Measures in Text Classification. Neural Computing and Applications, 35, 10521–10536.
Tang, L., & He, X. (2024). An Ensemble Approach to KNN with Multiple Distance Metrics. In Proceedings of the AAAI Conference on Artificial Intelligence.
Rahman, M., & Ahmad, M. (2020). Feature Selection in KNN for Medical Diagnosis. Journal of Biomedical Informatics, 103, 103383.
Sun, Y., & Zhang, W. (2021). Dynamic Weighting in KNN Classification. International Journal of Computer Vision, 129(1), 89–105.
Luo, J., & Wu, Q. (2022). Impact of Distance Metrics on KNN Performance in Image Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
Patel, D., & Singh, R. (2023). A Study on KNN Classification in Large Datasets. Neural Networks Journal, 156, 45–59.
Zhao, L., & Tan, J. (2024). Advancements in KNN-Based Recommender Systems. ACM Transactions on Knowledge Discovery from Data (TKDD), 18(2), 1–25.
