A new k-harmonic nearest neighbor classifier based on the multi-local means

作者：Pan, Zhibin^*; Wang, Yidi; Ku, Weiping

来源：Expert Systems with Applications, 2017, 67: 115-125.

DOI：10.1016/j.eswa.2016.09.031

• Local mean
• k-nearest neighbor
• Harmonic mean distance
• Pattern classification
• Small training sample size

The k-nearest neighbor (KNN) rule is a classical and yet very effective nonparametric technique in pattern classification, but its classification performance severely relies on the outliers. The local mean-based k-nearest neighbor classifier (LMKNN) was firstly introduced to achieve robustness against outliers by computing the local mean vector of k nearest neighbors for each class. However, its performances suffer from the choice of the single value of k for each class and the uniform value of k for different classes. In this paper, we propose a new KNN-based classifier, called multi-local means-based k-harmonic nearest neighbor (MLM-KHNN) rule. In our method, the k nearest neighbors in each class are first found, and then used to compute k different local mean vectors, which are employed to compute their harmonic mean distance to the query sample. Finally, MLM-KHNN proceeds in, classifying the query sample to the class with the minimum harmonic mean distance. The experimental results, based on twenty real-world datasets from UCI and KEEL repository, demonstrated that the proposed MLM-KHNN classifier achieves lower classification error rate and is less sensitive to the parameter k, when compared to nine related competitive KNN-based classifiers, especially in small training sample size situations.

• 出版日期2017-1
• 单位西安交通大学