Fuzzy K-Nearest Neighbor to Predict Rainfall in Padang Pariaman District

Penulis

  • Annisa Rizki Amalia Universitas Negeri Padang
  • Nonong Amalita
  • Yenni Kurniawati
  • Zamahsary Martha

DOI:

https://doi.org/10.24036/ujsds/vol2-iss1/126

Kata Kunci:

Classification, Data Mining, Fuzzy K-Nearest Neighbor, Rainfall

Abstrak

Rainfall is the amount of water that falls to the ground surface during a certain period which is measured in millimeters. The amount of rainfall can be estimated or predicted. One method used to predict rainfall is Data Mining, namely computer learning to analyze knowledge automatically so that a perfect new model is obtained. One of the best prediction algorithms in data mining is Fuzzy K-Nearest Neighbor (FK-NN) which uses the largest membership degree value from the test data in each class to determine the class. The number of sample classes obtained from rainfall data in Padang Pariaman Regency experienced unbalanced classes. One way to handle imbalance class cases is to use the Synthetic Minority Over-sampling Technique (SMOTE) technique which produces as much minority data as majority data. The results obtained in this study used the FK-NN classification with a total of 343 test data, parameter K=12 and using the Euclidean distance. The accuracy value was quite good, namely 76,38%.

Unduhan

Diterbitkan

2024-02-25

Cara Mengutip

Rizki Amalia, A., Nonong Amalita, Yenni Kurniawati, & Zamahsary Martha. (2024). Fuzzy K-Nearest Neighbor to Predict Rainfall in Padang Pariaman District. UNP Journal of Statistics and Data Science, 2(1), 64–70. https://doi.org/10.24036/ujsds/vol2-iss1/126

Artikel paling banyak dibaca berdasarkan penulis yang sama

1 2 3 4 5 6 > >>