An Attribute Contribution-Based K-Nearest Neighbor Classifier

K-nearest neighbor algorithm (KNN) is one of the most representative methods in data mining classification techniques. However, the KNN algorithm has a problem that when the traditional Euclidean distance formula is used to calculate the nearest neighbor distance, we ignore the relationship between attributes in the feature space. To tackle this issue, a covariance matrix is used to calculate the attribute contribution of the samples in order to solve the above problem. So an attribute contribution-based k-nearest neighbor classifier (ACWKNN) is proposed in this paper. The proposed algorithm is compared and experimented on the UCI standard dataset, and the results show that the method outperforms other KNN algorithms.