Achmad Ridok, Wayan Firdaus Mahmudy, Muhaimin Rifai
Learning algorithm is faced with the problem of selecting an optimized subset of original features to get the best performance. To do this, feature selection algorithm can be used to select a subset from the original features. Feature selection algorithm conducts this selection by removing irrelevant or redundant features from the chosen data set. The best features will improve the performance of the classification algorithm, particularly in accuracy and efficiency. This paper proposes fast correlation based filter (FCBF) methods of selection feature as preprocessing data and AIRS2 algorithm as a model prediction. Experimentation of the proposed method by 10 fold cross validation test on the classification of the dataset Breast-Cancer-Wisconsin reach correctness 100% for all setting k on K-Nearest Neighbors (KNN) from 1 until 30. © 2017 IEEE.
Filkom, Brawijaya University, Malang, Indonesia; Biologi FMIPA, Brawijaya University, Malang, Indonesia
