Attribute reduction is one of the commonly used techniques in data analysis and modeling. The granular-ball neighborhood rough set, which can adaptively set the neighborhood radius, enhances the accuracy and robustness of attribute reduction. However, current granular-ball generation methods face problems of uncertain numbers and unstable distributions. To address this issue, this paper proposed a granular-ball generation method based on density peaks. By using density peak points and the nearest centroid points as centers, this method ensures that the centers are composed of sample points, thereby enhancing the interpretability of granular-balls. Based on this new granular-ball generation method, a granular-ball neighborhood rough set model based on density peaks was derived. This model overcomes the limitation of using the positive region for attribute reduction in granular-ball neighborhood rough sets. And accordingly a backward attribute reduction algorithm was designed. The above algorithm was tested on multiple datasets. Experimental results show that, compared to existing methods, the new model achieves stable performance during the granular-ball generation process, and the reduced attributes significantly enhance classification performance.
HUQ H, YUD R, XIEZ X. Numerical Attribute Reduction Based on Neighborhood Granulation and Rough Approximation[J]. J Softw, 2008, 19(3): 640-649. DOI: 10.3724/SP.J.1001.2008.00640 .
[6]
HUQ H, YUD R, LIUJ F, et al. Neighborhood Rough Set Based Heterogeneous Feature Subset Selection[J]. Inf Sci, 2008, 178(18): 3577-3594. DOI:10.1016/j.ins.2008.05.024 .
[7]
WANGC Z, SHAOM W, HEQ, et al. Feature Subset Selection Based on Fuzzy Neighborhood Rough Sets[J]. Knowl Based Syst, 2016, 111: 173-179. DOI:10.1016/j.knosys.2016.08.009 .
[8]
XUW H, YUANZ T, LIUZ. Feature Selection for Unbalanced Distribution Hybrid Data Based on K-nearest Neighborhood Rough Set[J]. IEEE Trans Artif Intell, 2024, 5(1): 229-243. DOI:10.1109/TAI.2023.3237203 .
HUQ H, ZHAOH, YUD R. Efficient Symbolic and Numerical Attribute Reduction with Neighborhood Rough Sets[J]. Pattern Recognit Artif Intell, 2008, 21(6): 732-738. DOI: 10.3969/j.issn.1003-6059.2008.06.004 .
[11]
HUQ H, PEDRYCZW, YUD R, et al. Selecting Discrete and Continuous Features Based on Neighborhood Decision Error Minimization[J]. IEEE Trans Syst Man Cybern B Cybern, 2010, 40(1): 137-150. DOI:10.1109/TSMCB.2009.2024166 .
[12]
CHENH M, LIT R, FANX, et al. Feature Selection for Imbalanced Data Based on Neighborhood Rough Sets[J]. Inf Sci, 2019, 483: 1-20. DOI:10.1016/j.ins.2019.01.041 .
DUANJ, HUQ H, ZHANGL J, et al. Feature Selection for Multi-label Classification Based on Neighborhood Rough Sets[J]. J Comput Res Dev, 2015, 52(1): 56-65. DOI: 10.7544/issn.1000-1239.2015.20140544 .
[15]
HUM, TSANGE C C, GUOY T, et al. A Novel Approach to Attribute Reduction Based on Weighted Neighborhood Rough Sets[J]. Knowl Based Syst, 2021, 220: 106908. DOI:10.1016/j.knosys.2021.106908 .
[16]
XIAS Y, ZHANGH, LIW H, et al. GBNRS: a Novel Rough Set Algorithm for Fast Adaptive Attribute Reduction in Classification[J]. IEEE Trans Knowl Data Eng, 2022, 34(3): 1231-1242. DOI:10.1109/TKDE.2020.2997039 .
[17]
XIAS Y, WANGC, WANGG Y, et al. GBRS: a Unified Granular-ball Learning Model of Pawlak Rough Set and Neighborhood Rough Set[J]. IEEE Trans Neural Netw Learn Syst, 2025, 36(1): 1719-1733. DOI:10.1109/TNNLS.2023.3325199 .
[18]
XIAS Y, LIUY S, DINGX, et al. Granular Ball Computing Classifiers for Efficient, Scalable and Robust Learning[J]. Inf Sci, 2019, 483: 136-152. DOI:10.1016/j.ins.2019.01.010 .
[19]
XIEJ, KONGW Y, XIAS Y, et al. An Efficient Spectral Clustering Algorithm Based on Granular-ball[J]. IEEE Trans Knowl Data Eng, 2023, 35(9): 9743-9753. DOI:10.1109/TKDE.2023.3249475 .
[20]
XIAS Y, PENGD W, MENGD Y, et al. Ball k-Means: Fast Adaptive Clustering With No Bounds[J]. IEEE Trans Pattern Anal Mach Intell, 2022, 44(1): 87-99. DOI:10.1109/tpami.2020.3008694 .
[21]
CHENGD D, LIY, XIAS Y, et al. A Fast Granular-ball-based Density Peaks Clustering Algorithm for Large-scale Data[J]. IEEE Trans Neural Netw Learn Syst, 2024, 35(12): 17202-17215. DOI:10.1109/TNNLS.2023.3300916 .
[22]
XIAS Y, ZHENGS Y, WANGG Y, et al. Granular Ball Sampling for Noisy Label Classification or Imbalanced Classification[J]. IEEE Trans Neural Netw Learn Syst, 2023, 34(4): 2144-2155. DOI:10.1109/TNNLS.2021.3105984 .
[23]
QIANW B, XUF K, HUANGJ T, et al. A Novel Granular Ball Computing-based Fuzzy Rough Set for Feature Selection in Label Distribution Learning[J]. Knowl Based Syst, 2023, 278: 110898. DOI:10.1016/j.knosys.2023.110898 .
[24]
RODRIGUEZA, LAIOA. Clustering by Fast Search and Find of Density Peaks[J]. Science, 2014, 344(6191): 1492-1496. DOI:10.1126/science.1242072 .
[25]
XIAS Y, DAIX C, WANGG Y, et al. An Efficient and Adaptive Granular-ball Generation Method in Classification Problem[J]. IEEE Trans Neural Netw Learn Syst, 2022, 35(4): 5319-5331. DOI:10.1109/TNNLS.2022.3203381 .
[26]
FANGY, CAOX M, WANGX, et al. Hypersphere Neighborhood Rough Set for Rapid Attribute Reduction[C]//Pacific-Asia Conference on Knowledge Discovery and Data Mining. Cham: Springer International Publishing, 2022: 161-173. DOI:10.1007/978-3-031-05936-0_13 .