PENANGANAN DATA KETIDAKSEIMBANGAN DALAM PENDEKATAN SMOTE GUNA MENINGKATKAN AKURASI ALGORITMA K-NN
Keywords:
Accuracy, Imbalance Data, Classification, K-Nearest Neighbor, SmoteAbstract
Classification of unbalanced data is a frequent problem in the field of machine learning and data mining. In this research, SMOTE (Synthetic Minority Oversampling Technique) is applied to solve the problem of class imbalance in the dataset. It focuses on K-Nearest Neighbors (K-NN) algorithm and analyzes the accuracy improvement after SMOTE implementation. The data recorded was 8,545 data, the number of attributes was 6 attributes and the number of classes was 2 classes. The results of this study show that the accuracy with SMOTE technique increases compared to without using SMOTE, for example with K = 11 the accuracy of the K-NN algorithm with SMOTE technique is 0.8741 higher than the accuracy of the K-NN algorithm without SMOTE of 0.8683. This shows that the use of SMOTE can be an effective solution to improve the accuracy of the K-NN algorithm on unbalanced datasets. Therefore, this study concludes that the application of SMOTE can improve the accuracy of the K-NN algorithm on unbalanced data.References
Chao-Ren Wang and Xin-Xue Zhao, "An Improving Majority Weighted Minority Oversampling technique for Imbalanced Classification Problem." IEEE Access, vol. 8, no. 1, pp. 14773-14783, 2020.
Kasanah, A. N., Muladi, M., and Pujianto, U., "Penerapan Teknik SMOTE untuk Mengatasi Imbalance Class dalam Klasifikasi Objektivitas Berita Online Menggunakan Algoritma KNN," Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi), vol. 3, no. 2, pp. 196-201, 2019.
R. Siringoringo, "Klasifikasi data tidak Seimbang menggunakan algoritma SMOTE dan k-nearest neighbor," Journal Information System Development (ISD), vol. 3, no. 1, 2018.
Siagian, N. A. (2021). Analisis Perbandingan Akurasi dalam Mengidentifikasi Jenis Kaca. *InfoTekJar: Jurnal Nasional Informatika dan Teknologi Jaringan*, 5(2), 283-294.
