A Novel Neighborhood‐Weighted Sampling Method for Imbalanced Datasets

The weighted sampling methods based on k-nearest neighbors have been demonstrated to be effective in solving the class imbalance problem. However, they usually ignore positional relationship between a sample and heterogeneous samples its neighborhood when calculating weight. This paper proposes novel neighborhood-weighted method named NWBBagging improve Bagging algorithm's performance imbalanced datasets. It considers center identifying critical samples. And parameter reduction is proposed combined into ensemble learning framework, which reduces parameters increases classifier's diversity. We compare with some state-of-the-art algorithms 34 datasets, result shows that achieves better performance.