Journal of Computer Applications ›› 2022, Vol. 42 ›› Issue (12): 3740-3749.DOI: 10.11772/j.issn.1001-9081.2021101756
Special Issue: 数据科学与技术
• Data science and technology • Previous Articles Next Articles
Jiaojiao GUAN(), Xuezhong QIAN, Shibing ZHOU, Kaibin JIANG, Wei SONG
Received:
2021-10-12
Revised:
2022-01-10
Accepted:
2022-01-24
Online:
2022-03-04
Published:
2022-12-10
Contact:
Jiaojiao GUAN
About author:
QIAN Xuezhong,born in 1967, M. S., associate professor. His research interests include data mining, machine learning, artificial intelligence.Supported by:
通讯作者:
管娇娇
作者简介:
钱雪忠(1967—),男,江苏无锡人,副教授,硕士,CCF会员,主要研究方向:数据挖掘、机器学习、人工智能基金资助:
CLC Number:
Jiaojiao GUAN, Xuezhong QIAN, Shibing ZHOU, Kaibin JIANG, Wei SONG. Multi-view clustering via subspace merging on Grassmann manifold[J]. Journal of Computer Applications, 2022, 42(12): 3740-3749.
管娇娇, 钱雪忠, 周世兵, 姜凯彬, 宋威. 基于格拉斯曼流形子空间融合的多视图聚类[J]. 《计算机应用》唯一官方网站, 2022, 42(12): 3740-3749.
Add to citation manager EndNote|Ris|BibTeX
URL: https://www.joca.cn/EN/10.11772/j.issn.1001-9081.2021101756
符号 | 含义 |
---|---|
数据点的个数 | |
聚类的个数 | |
视图索引 | |
视图的数目 | |
第 | |
第 | |
第 | |
第 | |
一致性亲和矩阵 | |
第 | |
一致性亲和矩阵的拉普拉斯矩阵 | |
第 |
Tab. 1 Common symbols and their definitions
符号 | 含义 |
---|---|
数据点的个数 | |
聚类的个数 | |
视图索引 | |
视图的数目 | |
第 | |
第 | |
第 | |
第 | |
一致性亲和矩阵 | |
第 | |
一致性亲和矩阵的拉普拉斯矩阵 | |
第 |
数据集 | 样本数 | 视图数 | 类别数 | #d1 | #d2 | #d3 | #d4 | #d5 | #d6 |
---|---|---|---|---|---|---|---|---|---|
MSRCV1 | 210 | 6 | 7 | 1 302 | 48 | 512 | 100 | 256 | 210 |
Yale | 165 | 3 | 15 | 3 096 | 3 304 | 6 750 | |||
Prokaryotic | 551 | 3 | 4 | 393 | 3 | 438 | |||
HW2sources | 2 000 | 2 | 10 | 784 | 256 | ||||
Not-Hill | 550 | 3 | 5 | 2 000 | 3 304 | 6 750 | |||
WebKB | 203 | 3 | 4 | 1 703 | 230 | 230 |
Tab.2 Experimental datasets
数据集 | 样本数 | 视图数 | 类别数 | #d1 | #d2 | #d3 | #d4 | #d5 | #d6 |
---|---|---|---|---|---|---|---|---|---|
MSRCV1 | 210 | 6 | 7 | 1 302 | 48 | 512 | 100 | 256 | 210 |
Yale | 165 | 3 | 15 | 3 096 | 3 304 | 6 750 | |||
Prokaryotic | 551 | 3 | 4 | 393 | 3 | 438 | |||
HW2sources | 2 000 | 2 | 10 | 784 | 256 | ||||
Not-Hill | 550 | 3 | 5 | 2 000 | 3 304 | 6 750 | |||
WebKB | 203 | 3 | 4 | 1 703 | 230 | 230 |
算法 | 参数 |
---|---|
Co-reg | |
MLRSSC | |
LMSC | |
DiMSC | |
ECMSC | |
DALIGA |
Tab. 3 Parameters setting of algorithms to be compared
算法 | 参数 |
---|---|
Co-reg | |
MLRSSC | |
LMSC | |
DiMSC | |
ECMSC | |
DALIGA |
数据集 | 算法 | NMI | ACC | ARI | F | P | R |
---|---|---|---|---|---|---|---|
MSRC V1 | SCbest | 0.628 9(0.048 9) | 0.721 7(0.066 5) | 0.535 3(0.067 9) | 0.600 5(0.057 9) | 0.594 8(0.060 6) | 0.606 4(0.055 2) |
Co-reg | 0.677 4(0.015 1) | 0.759 8(0.018 0) | 0.600 9(0.020 1) | 0.658 6(0.017 1) | 0.635 2(0.018 1) | 0.684 5(0.016 4) | |
Co-train | 0.746 3(0.039 1) | 0.836 0(0.055 6) | 0.689 5(0.054 6) | 0.733 2(0.046 3) | 0.725 6(0.054 0) | 0.741 1(0.038 2) | |
MLRSSC | 0.538 2(0.024 2) | 0.686 0(0.030 0) | 0.450 0(0.034 9) | 0.528 2(0.029 6) | 0.515 3(0.031 0) | 0.541 9(0.028 0) | |
KMLRSSC | 0.666 0(0.037 1) | 0.740 5(0.070 3) | 0.586 5(0.060 5) | 0.645 5(0.050 8) | 0.629 6(0.057 5) | 0.662 7(0.043 6) | |
MLAN | 0.661 6(0.000 0) | 0.681 0(0.000 0) | 0.503 9(0.000 0) | 0.585 5(0.000 0) | 0.494 1(0.000 0) | 0.718 2(0.000 0) | |
LMSC | 0.734 5(0.000 0) | 0.694 0(0.000 0) | 0.737 0(0.000 0) | 0.726 9(0.000 0) | 0.747 5(0.000 0) | ||
DiMSC | 0.697 9(0.003 6) | 0.801 2(0.005 8) | 0.621 8(0.007 0) | 0.675 4(0.005 7) | 0.661 0(0.008 7) | 0.690 6(0.002 7) | |
ECMSC | 0.841 0(0.057 1) | ||||||
DALIGA | 0.702 8(0.010 7) | 0.810 9(0.008 2) | 0.639 4(0.013 0) | 0.690 3(0.011 0) | 0.678 6(0.013 0) | 0.702 4(0.009 2) | |
本文方法 | 0.874 3(0.000 0) | 0.928 6(0.000 0) | 0.835 5(0.000 0) | 0.858 6(0.000 0) | 0.849 1(0.000 0) | 0.868 3(0.000 0) | |
Yale | SCbest | 0.639 0(0.032 2) | 0.594 2(0.043 9) | 0.414 1(0.044 3) | 0.451 6(0.041 1) | 0.431 2(0.043 6) | 0.474 4(0.039 9) |
Co-reg | 0.657 6(0.011 5) | 0.607 9(0.015 5) | 0.443 5(0.016 2) | 0.479 3(0.015 1) | 0.455 1(0.015 0) | 0.506 7(0.015 6) | |
Co-train | 0.678 3(0.027 1) | 0.640 0(0.044 8) | 0.467 5(0.039 0) | 0.501 4(0.036 5) | 0.481 1(0.036 5) | 0.523 8(0.038 3) | |
MLRSSC | 0.686 4(0.031 6) | 0.637 6(0.050 0) | 0.461 5(0.049 4) | 0.499 5(0.046 0) | 0.467 2(0.048 4) | 0.530 4(0.046 4) | |
KMLRSSC | 0.688 6(0.046 5) | 0.643 3(0.064 5) | 0.486 0(0.068 0) | 0.519 0(0.063 3) | 0.494 1(0.065 1) | 0.546 8(0.061 7) | |
MLAN | 0.690 8(0.001 4) | 0.641 8(0.001 9) | 0.475 8(0.001 0) | 0.510 4(0.001 0) | 0.472 5(0.000 2) | 0.554 8(0.002 6) | |
LMSC | 0.696 1(0.000 0) | 0.660 6(0.000 0) | 0.458 5(0.000 0) | 0.494 5(0.000 0) | 0.453 1(0.000 0) | 0.544 2(0.000 0) | |
DiMSC | 0.710 5(0.022 5) | 0.551 3(0.029 8) | 0.531 8(0.030 3) | 0.572 4(0.029 6) | |||
ECMSC | 0.680 3(0.018 2) | 0.637 1(0.040 5) | 0.416 7(0.023 8) | 0.456 4(0.021 4) | 0.409 7(0.028 9) | 0.516 7(0.021 2) | |
DALIGA | 0.722 4(0.000 0) | 0.690 9(0.000 0) | 0.520 8(0.000 0) | ||||
本文方法 | 0.703 0(0.000 0) | 0.545 2(0.000 0) | 0.573 8(0.000 0) | 0.557 1(0.000 0) | 0.591 5(0.000 0) | ||
Prokaryotic | SCbest | 0.378 5(0.070 8) | 0.592 6(0.095 6) | 0.290 5(0.098 5) | 0.521 8(0.070 8) | 0.629 8(0.080 1) | 0.446 1(0.066 2) |
Co-reg | 0.327 4(0.011 0) | 0.550 9(0.017 6) | 0.222 3(0.020 6) | 0.476 1(0.014 1) | 0.575 6(0.017 1) | 0.406 5(0.012 5) | |
Co-train | 0.516 1(0.041 3) | ||||||
MLRSSC | 0.324 4(0.009 0) | 0.660 1(0.005 3) | 0.350 9(0.026 4) | 0.598 6(0.027 5) | 0.617 6(0.007 6) | 0.583 5(0.059 1) | |
KMLRSSC | 0.406 8(0.036 8) | 0.655 0(0.063 5) | 0.359 5(0.075 4) | 0.562 2(0.053 7) | 0.698 7(0.057 9) | 0.470 5(0.049 6) | |
MLAN | 0.239 3(0.000 0) | 0.626 1(0.000 0) | 0.131 8(0.000 0) | 0.577 4(0.000 0) | 0.443 5(0.000 0) | 0.8270(0.0000) | |
LMSC | 0.184 5(0.000 0) | 0.437 4(0.000 0) | 0.078 3(0.000 0) | 0.383 8(0.000 0) | 0.456 5(0.000 0) | 0.331 0(0.000 0) | |
DiMSC | 0.057 6(0.000 0) | 0.390 2(0.000 0) | 0.042 0(0.000 0) | 0.346 3(0.000 1) | 0.429 8(0.000 0) | 0.290 0(0.000 0) | |
ECMSC | 0.195 3(0.011 3) | 0.437 9(0.015 6) | 0.082 5(0.010 0) | 0.381 1(0.007 3) | 0.461 7(0.008 4) | 0.324 5(0.007 8) | |
DALIGA | 0.050 7(0.000 0) | 0.401 1(0.000 0) | 0.039 6(0.000 0) | 0.403 4(0.000 0) | 0.419 8(0.000 0) | 0.388 3(0.000 0) | |
本文方法 | 0.5015(0.0000) | 0.7731(0.0000) | 0.5115(0.0000) | 0.6779(0.0000) | 0.7848(0.0000) | ||
HW2sources | SCbest | 0.461 2(0.012 5) | 0.543 1(0.029 9) | 0.356 5(0.018 6) | 0.421 7(0.016 6) | 0.414 1(0.016 7) | 0.429 6(0.016 7) |
Co-reg | 0.694 6(0.024 0) | 0.780 3(0.028 4) | 0.636 9(0.031 6) | 0.673 5(0.028 3) | 0.666 2(0.028 9) | 0.681 0(0.027 8) | |
Co-train | 0.822 2(0.014 2) | 0.810 4(0.026 6) | 0.829 4(0.023 7) | 0.827 1(0.030 0) | 0.831 7(0.016 8) | ||
MLRSSC | 0.490 2(0.011 9) | 0.568 6(0.030 9) | 0.386 7(0.016 4) | 0.450 4(0.014 3) | 0.432 0(0.017 8) | 0.470 9(0.015 8) | |
KMLRSSC | 0.685 2(0.023 2) | 0.733 4(0.044 1) | 0.604 5(0.031 4) | 0.644 5(0.027 9) | 0.635 9(0.033 9) | 0.653 6(0.024 1) | |
MLAN | 0.888 8(0.000 5) | 0.835 4(0.000 3) | 0.814 7(0.000 6) | 0.834 5(0.000 5) | 0.778 7(0.000 6) | 0.898 8(0.000 4) | |
LMSC | 0.696 5(0.000 0) | 0.779 0(0.000 0) | 0.631 4(0.000 0) | 0.668 2(0.000 0) | 0.666 2(0.000 0) | 0.670 3(0.000 0) | |
DiMSC | 0.698 1(0.000 9) | 0.761 5(0.000 7) | 0.623 3(0.000 8) | 0.661 0(0.000 7) | 0.657 0(0.000 8) | 0.665 2(0.000 6) | |
ECMSC | 0.904 0(0.066 0) | ||||||
DALIGA | 0.466 5(0.000 0) | 0.522 0(0.000 0) | 0.371 9(0.000 0) | 0.446 1(0.000 0) | 0.376 3(0.000 0) | 0.547 6(0.000 0) | |
本文方法 | 0.9635(0.0000) | 0.9835(0.0000) | 0.9640(0.0000) | 0.9676(0.0000) | 0.9676(0.0000) | 0.9677(0.0000) | |
Not-Hill | SCbest | 0.712 5(0.032 3) | 0.841 6(0.057 4) | 0.709 4(0.067 8) | 0.773 7(0.048 7) | 0.771 7(0.076 2) | 0.778 6(0.016 7) |
Co-reg | 0.747 3(0.018 7) | 0.793 9(0.013 7) | 0.713 5(0.023 4) | 0.775 8(0.018 4) | 0.774 8(0.017 3) | 0.777 1(0.019 7) | |
Co-train | 0.777 2(0.005 2) | 0.828 5(0.004 2) | 0.764 3(0.004 7) | 0.816 4(0.003 7) | 0.803 0(0.002 8) | 0.830 3(0.004 8) | |
MLRSSC | 0.655 0(0.011 0) | 0.693 1(0.007 1) | 0.529 9(0.018 2) | 0.634 8(0.011 6) | 0.620 6(0.027 9) | 0.650 6(0.006 2) | |
KMLRSSC | 0.795 4(0.023 9) | 0.823 9(0.023 2) | 0.774 7(0.040 7) | 0.823 9(0.032 0) | 0.820 4(0.031 7) | 0.827 6(0.033 6) | |
MLAN | 0.724 0(0.000 0) | 0.678 2(0.000 0) | 0.570 0(0.000 0) | 0.684 2(0.000 0) | 0.560 7(0.000 0) | 0.877 5(0.000 0) | |
LMSC | 0.655 2(0.000 0) | 0.734 5(0.000 0) | 0.626 3(0.000 0) | 0.707 0(0.000 0) | 0.712 6(0.000 0) | 0.701 5(0.000 0) | |
DiMSC | 0.793 7(0.010 8) | 0.840 9(0.006 0) | 0.781 5(0.013 9) | 0.829 6(0.011 0) | 0.819 6(0.008 8) | 0.840 0(0.013 3) | |
ECMSC | 0.759 8(0.051 5) | 0.809 6(0.050 3) | 0.632 5(0.081 2) | 0.715 0(0.060 5) | 0.698 1(0.075 2) | 0.733 6(0.043 6) | |
DALIGA | 0.9088(0.0000) | ||||||
本文方法 | 0.8687(0.0000) | 0.9346(0.0000) | 0.8941(0.0000) | 0.9175(0.0000) | 0.9287(0.0000) | ||
WebKB | SCbest | 0.321 4(0.040 0) | 0.570 4(0.024 9) | 0.320 8(0.055 0) | 0.548 7(0.043 9) | 0.640 4(0.026 4) | 0.481 1(0.055 7) |
Co-reg | 0.406 1(0.010 7) | 0.607 3(0.008 7) | 0.370 9(0.011 0) | 0.571 9(0.007 8) | 0.703 3(0.009 2) | 0.482 5(0.008 5) | |
Co-train | 0.418 0(0.010 8) | 0.655 4(0.036 4) | 0.429 1(0.020 5) | 0.629 4(0.015 6) | 0.568 6(0.023 1) | ||
MLRSSC | 0.679 1(0.017 1) | 0.662 7(0.004 6) | 0.782 3(0.041 4) | ||||
KMLRSSC | 0.461 3(0.021 5) | 0.662 6(0.025 0) | 0.438 4(0.020 8) | 0.627 6(0.017 2) | 0.7327(0.0182) | 0.549 7(0.028 8) | |
MLAN | 0.402 2(0.000 0) | 0.729 1(0.000 0) | 0.373 2(0.000 0) | 0.667 9(0.000 0) | 0.558 5(0.000 0) | 0.830 7(0.000 0) | |
LMSC | N/A | N/A | N/A | N/A | N/A | N/A | |
DiMSC | 0.372 1(0.000 0) | 0.650 2(0.000 0) | 0.389 8(0.000 0) | 0.591 6(0.000 0) | 0.703 3(0.000 0) | 0.510 5(0.000 0) | |
ECMSC | N/A | N/A | N/A | N/A | N/A | N/A | |
DALIGA | 0.466 1(0.000 0) | 0.465 7(0.000 0) | 0.707 8(0.000 0) | 0.614 1(0.000 0) | |||
本文方法 | 0.5257(0.0000) | 0.8227(0.0000) | 0.6065(0.0000) | 0.7820(0.0000) | 0.689 5(0.000 0) | 0.9031(0.0000) |
Tab. 4 Clustering performance on MSRCV1, Yale, Prokaryotic, HW2sources, Not-Hill, WebKB datasets among different algorithms
数据集 | 算法 | NMI | ACC | ARI | F | P | R |
---|---|---|---|---|---|---|---|
MSRC V1 | SCbest | 0.628 9(0.048 9) | 0.721 7(0.066 5) | 0.535 3(0.067 9) | 0.600 5(0.057 9) | 0.594 8(0.060 6) | 0.606 4(0.055 2) |
Co-reg | 0.677 4(0.015 1) | 0.759 8(0.018 0) | 0.600 9(0.020 1) | 0.658 6(0.017 1) | 0.635 2(0.018 1) | 0.684 5(0.016 4) | |
Co-train | 0.746 3(0.039 1) | 0.836 0(0.055 6) | 0.689 5(0.054 6) | 0.733 2(0.046 3) | 0.725 6(0.054 0) | 0.741 1(0.038 2) | |
MLRSSC | 0.538 2(0.024 2) | 0.686 0(0.030 0) | 0.450 0(0.034 9) | 0.528 2(0.029 6) | 0.515 3(0.031 0) | 0.541 9(0.028 0) | |
KMLRSSC | 0.666 0(0.037 1) | 0.740 5(0.070 3) | 0.586 5(0.060 5) | 0.645 5(0.050 8) | 0.629 6(0.057 5) | 0.662 7(0.043 6) | |
MLAN | 0.661 6(0.000 0) | 0.681 0(0.000 0) | 0.503 9(0.000 0) | 0.585 5(0.000 0) | 0.494 1(0.000 0) | 0.718 2(0.000 0) | |
LMSC | 0.734 5(0.000 0) | 0.694 0(0.000 0) | 0.737 0(0.000 0) | 0.726 9(0.000 0) | 0.747 5(0.000 0) | ||
DiMSC | 0.697 9(0.003 6) | 0.801 2(0.005 8) | 0.621 8(0.007 0) | 0.675 4(0.005 7) | 0.661 0(0.008 7) | 0.690 6(0.002 7) | |
ECMSC | 0.841 0(0.057 1) | ||||||
DALIGA | 0.702 8(0.010 7) | 0.810 9(0.008 2) | 0.639 4(0.013 0) | 0.690 3(0.011 0) | 0.678 6(0.013 0) | 0.702 4(0.009 2) | |
本文方法 | 0.874 3(0.000 0) | 0.928 6(0.000 0) | 0.835 5(0.000 0) | 0.858 6(0.000 0) | 0.849 1(0.000 0) | 0.868 3(0.000 0) | |
Yale | SCbest | 0.639 0(0.032 2) | 0.594 2(0.043 9) | 0.414 1(0.044 3) | 0.451 6(0.041 1) | 0.431 2(0.043 6) | 0.474 4(0.039 9) |
Co-reg | 0.657 6(0.011 5) | 0.607 9(0.015 5) | 0.443 5(0.016 2) | 0.479 3(0.015 1) | 0.455 1(0.015 0) | 0.506 7(0.015 6) | |
Co-train | 0.678 3(0.027 1) | 0.640 0(0.044 8) | 0.467 5(0.039 0) | 0.501 4(0.036 5) | 0.481 1(0.036 5) | 0.523 8(0.038 3) | |
MLRSSC | 0.686 4(0.031 6) | 0.637 6(0.050 0) | 0.461 5(0.049 4) | 0.499 5(0.046 0) | 0.467 2(0.048 4) | 0.530 4(0.046 4) | |
KMLRSSC | 0.688 6(0.046 5) | 0.643 3(0.064 5) | 0.486 0(0.068 0) | 0.519 0(0.063 3) | 0.494 1(0.065 1) | 0.546 8(0.061 7) | |
MLAN | 0.690 8(0.001 4) | 0.641 8(0.001 9) | 0.475 8(0.001 0) | 0.510 4(0.001 0) | 0.472 5(0.000 2) | 0.554 8(0.002 6) | |
LMSC | 0.696 1(0.000 0) | 0.660 6(0.000 0) | 0.458 5(0.000 0) | 0.494 5(0.000 0) | 0.453 1(0.000 0) | 0.544 2(0.000 0) | |
DiMSC | 0.710 5(0.022 5) | 0.551 3(0.029 8) | 0.531 8(0.030 3) | 0.572 4(0.029 6) | |||
ECMSC | 0.680 3(0.018 2) | 0.637 1(0.040 5) | 0.416 7(0.023 8) | 0.456 4(0.021 4) | 0.409 7(0.028 9) | 0.516 7(0.021 2) | |
DALIGA | 0.722 4(0.000 0) | 0.690 9(0.000 0) | 0.520 8(0.000 0) | ||||
本文方法 | 0.703 0(0.000 0) | 0.545 2(0.000 0) | 0.573 8(0.000 0) | 0.557 1(0.000 0) | 0.591 5(0.000 0) | ||
Prokaryotic | SCbest | 0.378 5(0.070 8) | 0.592 6(0.095 6) | 0.290 5(0.098 5) | 0.521 8(0.070 8) | 0.629 8(0.080 1) | 0.446 1(0.066 2) |
Co-reg | 0.327 4(0.011 0) | 0.550 9(0.017 6) | 0.222 3(0.020 6) | 0.476 1(0.014 1) | 0.575 6(0.017 1) | 0.406 5(0.012 5) | |
Co-train | 0.516 1(0.041 3) | ||||||
MLRSSC | 0.324 4(0.009 0) | 0.660 1(0.005 3) | 0.350 9(0.026 4) | 0.598 6(0.027 5) | 0.617 6(0.007 6) | 0.583 5(0.059 1) | |
KMLRSSC | 0.406 8(0.036 8) | 0.655 0(0.063 5) | 0.359 5(0.075 4) | 0.562 2(0.053 7) | 0.698 7(0.057 9) | 0.470 5(0.049 6) | |
MLAN | 0.239 3(0.000 0) | 0.626 1(0.000 0) | 0.131 8(0.000 0) | 0.577 4(0.000 0) | 0.443 5(0.000 0) | 0.8270(0.0000) | |
LMSC | 0.184 5(0.000 0) | 0.437 4(0.000 0) | 0.078 3(0.000 0) | 0.383 8(0.000 0) | 0.456 5(0.000 0) | 0.331 0(0.000 0) | |
DiMSC | 0.057 6(0.000 0) | 0.390 2(0.000 0) | 0.042 0(0.000 0) | 0.346 3(0.000 1) | 0.429 8(0.000 0) | 0.290 0(0.000 0) | |
ECMSC | 0.195 3(0.011 3) | 0.437 9(0.015 6) | 0.082 5(0.010 0) | 0.381 1(0.007 3) | 0.461 7(0.008 4) | 0.324 5(0.007 8) | |
DALIGA | 0.050 7(0.000 0) | 0.401 1(0.000 0) | 0.039 6(0.000 0) | 0.403 4(0.000 0) | 0.419 8(0.000 0) | 0.388 3(0.000 0) | |
本文方法 | 0.5015(0.0000) | 0.7731(0.0000) | 0.5115(0.0000) | 0.6779(0.0000) | 0.7848(0.0000) | ||
HW2sources | SCbest | 0.461 2(0.012 5) | 0.543 1(0.029 9) | 0.356 5(0.018 6) | 0.421 7(0.016 6) | 0.414 1(0.016 7) | 0.429 6(0.016 7) |
Co-reg | 0.694 6(0.024 0) | 0.780 3(0.028 4) | 0.636 9(0.031 6) | 0.673 5(0.028 3) | 0.666 2(0.028 9) | 0.681 0(0.027 8) | |
Co-train | 0.822 2(0.014 2) | 0.810 4(0.026 6) | 0.829 4(0.023 7) | 0.827 1(0.030 0) | 0.831 7(0.016 8) | ||
MLRSSC | 0.490 2(0.011 9) | 0.568 6(0.030 9) | 0.386 7(0.016 4) | 0.450 4(0.014 3) | 0.432 0(0.017 8) | 0.470 9(0.015 8) | |
KMLRSSC | 0.685 2(0.023 2) | 0.733 4(0.044 1) | 0.604 5(0.031 4) | 0.644 5(0.027 9) | 0.635 9(0.033 9) | 0.653 6(0.024 1) | |
MLAN | 0.888 8(0.000 5) | 0.835 4(0.000 3) | 0.814 7(0.000 6) | 0.834 5(0.000 5) | 0.778 7(0.000 6) | 0.898 8(0.000 4) | |
LMSC | 0.696 5(0.000 0) | 0.779 0(0.000 0) | 0.631 4(0.000 0) | 0.668 2(0.000 0) | 0.666 2(0.000 0) | 0.670 3(0.000 0) | |
DiMSC | 0.698 1(0.000 9) | 0.761 5(0.000 7) | 0.623 3(0.000 8) | 0.661 0(0.000 7) | 0.657 0(0.000 8) | 0.665 2(0.000 6) | |
ECMSC | 0.904 0(0.066 0) | ||||||
DALIGA | 0.466 5(0.000 0) | 0.522 0(0.000 0) | 0.371 9(0.000 0) | 0.446 1(0.000 0) | 0.376 3(0.000 0) | 0.547 6(0.000 0) | |
本文方法 | 0.9635(0.0000) | 0.9835(0.0000) | 0.9640(0.0000) | 0.9676(0.0000) | 0.9676(0.0000) | 0.9677(0.0000) | |
Not-Hill | SCbest | 0.712 5(0.032 3) | 0.841 6(0.057 4) | 0.709 4(0.067 8) | 0.773 7(0.048 7) | 0.771 7(0.076 2) | 0.778 6(0.016 7) |
Co-reg | 0.747 3(0.018 7) | 0.793 9(0.013 7) | 0.713 5(0.023 4) | 0.775 8(0.018 4) | 0.774 8(0.017 3) | 0.777 1(0.019 7) | |
Co-train | 0.777 2(0.005 2) | 0.828 5(0.004 2) | 0.764 3(0.004 7) | 0.816 4(0.003 7) | 0.803 0(0.002 8) | 0.830 3(0.004 8) | |
MLRSSC | 0.655 0(0.011 0) | 0.693 1(0.007 1) | 0.529 9(0.018 2) | 0.634 8(0.011 6) | 0.620 6(0.027 9) | 0.650 6(0.006 2) | |
KMLRSSC | 0.795 4(0.023 9) | 0.823 9(0.023 2) | 0.774 7(0.040 7) | 0.823 9(0.032 0) | 0.820 4(0.031 7) | 0.827 6(0.033 6) | |
MLAN | 0.724 0(0.000 0) | 0.678 2(0.000 0) | 0.570 0(0.000 0) | 0.684 2(0.000 0) | 0.560 7(0.000 0) | 0.877 5(0.000 0) | |
LMSC | 0.655 2(0.000 0) | 0.734 5(0.000 0) | 0.626 3(0.000 0) | 0.707 0(0.000 0) | 0.712 6(0.000 0) | 0.701 5(0.000 0) | |
DiMSC | 0.793 7(0.010 8) | 0.840 9(0.006 0) | 0.781 5(0.013 9) | 0.829 6(0.011 0) | 0.819 6(0.008 8) | 0.840 0(0.013 3) | |
ECMSC | 0.759 8(0.051 5) | 0.809 6(0.050 3) | 0.632 5(0.081 2) | 0.715 0(0.060 5) | 0.698 1(0.075 2) | 0.733 6(0.043 6) | |
DALIGA | 0.9088(0.0000) | ||||||
本文方法 | 0.8687(0.0000) | 0.9346(0.0000) | 0.8941(0.0000) | 0.9175(0.0000) | 0.9287(0.0000) | ||
WebKB | SCbest | 0.321 4(0.040 0) | 0.570 4(0.024 9) | 0.320 8(0.055 0) | 0.548 7(0.043 9) | 0.640 4(0.026 4) | 0.481 1(0.055 7) |
Co-reg | 0.406 1(0.010 7) | 0.607 3(0.008 7) | 0.370 9(0.011 0) | 0.571 9(0.007 8) | 0.703 3(0.009 2) | 0.482 5(0.008 5) | |
Co-train | 0.418 0(0.010 8) | 0.655 4(0.036 4) | 0.429 1(0.020 5) | 0.629 4(0.015 6) | 0.568 6(0.023 1) | ||
MLRSSC | 0.679 1(0.017 1) | 0.662 7(0.004 6) | 0.782 3(0.041 4) | ||||
KMLRSSC | 0.461 3(0.021 5) | 0.662 6(0.025 0) | 0.438 4(0.020 8) | 0.627 6(0.017 2) | 0.7327(0.0182) | 0.549 7(0.028 8) | |
MLAN | 0.402 2(0.000 0) | 0.729 1(0.000 0) | 0.373 2(0.000 0) | 0.667 9(0.000 0) | 0.558 5(0.000 0) | 0.830 7(0.000 0) | |
LMSC | N/A | N/A | N/A | N/A | N/A | N/A | |
DiMSC | 0.372 1(0.000 0) | 0.650 2(0.000 0) | 0.389 8(0.000 0) | 0.591 6(0.000 0) | 0.703 3(0.000 0) | 0.510 5(0.000 0) | |
ECMSC | N/A | N/A | N/A | N/A | N/A | N/A | |
DALIGA | 0.466 1(0.000 0) | 0.465 7(0.000 0) | 0.707 8(0.000 0) | 0.614 1(0.000 0) | |||
本文方法 | 0.5257(0.0000) | 0.8227(0.0000) | 0.6065(0.0000) | 0.7820(0.0000) | 0.689 5(0.000 0) | 0.9031(0.0000) |
1 | KUMAR A, RAI P, DAUMÉ H, III. Co-regularized multi-view spectral clustering[C]// Proceedings of the 24th International Conference on Neural Information Processing Systems. Red Hook, NY: Curran Associates Inc., 2011:1413-1421. |
2 | KUMAR A, DAUMÉ H, III. A co-training approach for multi-view spectral clustering[C]// Proceedings of the 28th International Conference on International Conference on Machine Learning. Madison, WI: Omnipress, 2011:393-400. |
3 | WANG C D, CHEN M S, HUANG L, et al. Smoothness regularized multiview subspace clustering with kernel learning[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(11): 5047-5060. 10.1109/tnnls.2020.3026686 |
4 | ZHANG G Y, ZHOU Y R, HE X Y, et al. One-step kernel multi-view subspace clustering[J]. Knowledge-Based Systems, 2020, 189: No.105126. 10.1016/j.knosys.2019.105126 |
5 | ELHAMIFAR E, VIDAL R. Sparse subspace clustering[C]// Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2009:2790-2797. 10.1109/cvpr.2009.5206547 |
6 | LIU G C, LIN Z C, YAN S C, et al. Robust recovery of subspace structures by low-rank representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(1): 171-184. 10.1109/tpami.2012.88 |
7 | WANG Y X, XU H, LENG C L. Provable subspace clustering: when LRR meets SSC[C]// Proceedings of the 26th International Conference on Neural Information Processing Systems. Red Hook, NY: Curran Associates Inc., 2013:64-72. |
8 | GAO H C, NIE F P, LI X L, et al. Multi-view subspace clustering[C]// Proceedings of the 2015 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2015:4238-4246. 10.1109/iccv.2015.482 |
9 | BRBIĆ M, KOPRIVA I. Multi-view low-rank sparse subspace clustering[J]. Pattern Recognition, 2018, 73:247-258. 10.1016/j.patcog.2017.08.024 |
10 | ZHANG C Q, HU Q H, FU H Z, et al. Latent multi-view subspace clustering[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017:4333-4341. 10.1109/cvpr.2017.461 |
11 | CAO X C, ZHANG C Q, FU H Z, et al. Diversity-induced multi-view subspace clustering[C]// Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2015:586-594. 10.1109/cvpr.2015.7298657 |
12 | WANG X B, GUO X J, LEI Z, et al. Exclusivity-consistency regularized multi-view subspace clustering[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017:1-9. 10.1109/cvpr.2017.8 |
13 | RONG W T, ZHUO E H, PENG H, et al. Learning a consensus affinity matrix for multi-view clustering via subspaces merging on Grassmann manifold[J]. Information Sciences, 2021, 547:68-87. 10.1016/j.ins.2020.07.059 |
14 | ZHANG C Q, FU H Z, LIU S, et al. Low-rank tensor constrained multiview subspace clustering[C]// Proceedings of the 2015 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2015:1582-1590. 10.1109/iccv.2015.185 |
15 | CHEN Y Y, XIAO X L, ZHOU Y C. Multi-view subspace clustering via simultaneously learning the representation tensor and affinity matrix[J]. Pattern Recognition, 2020, 106: No.107441. 10.1016/j.patcog.2020.107441 |
16 | CHEN Y Y, WANG S Q, PENG C, et al. Generalized nonconvex low-rank tensor approximation for multi-view subspace clustering[J]. IEEE Transactions on Image Processing, 2021, 30: 4022-4035. 10.1109/tip.2021.3068646 |
17 | XIA W, ZHANG X D, GAO Q X, et al. Multiview subspace clustering by an enhanced tensor nuclear norm[J]. IEEE Transactions on Cybernetics, 2022, 52(9):8962-8975. 10.1109/tcyb.2021.3052352 |
18 | GUO J P, SUN Y F, GAO J B, et al. Rank consistency induced multiview subspace clustering via low-rank matrix factorization[J]. IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(7):3157-3170. 10.1109/tnnls.2021.3071797 |
19 | YIN M, LIU W, LI M S, et al. Cauchy loss induced block diagonal representation for robust multi-view subspace clustering[J]. Neurocomputing, 2021, 427:84-95. 10.1016/j.neucom.2020.11.017 |
20 | HU H, FENG J J, ZHOU J. Exploiting unsupervised and supervised constraints for subspace clustering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(8): 1542-1557. 10.1109/tpami.2014.2377740 |
21 | MOHAR B. The Laplacian spectrum of graphs[M]// ALVI Y, CHARTRAND G, OELLERMANN O R, et al. Graph Theory, Combinatorics and Applications: Proceedings of the Sixth Quadrennial International Conference on the Theory and Applications of Graphs, Vol. 2. Hoboken, NJ: John Wiley & Sons, Inc., 1991:871-898. |
22 | FAN K. On a theorem of Weyl concerning eigenvalues of linear transformations I[J]. Proceedings of the National Academy of Sciences of the United States of America, 1949, 35(11): 652-655. 10.1073/pnas.35.11.652 |
23 | NG A Y, JORDAN M I, WEISS Y. On spectral clustering: analysis and an algorithm[C]// Proceedings of the 14th International Conference on Neural Information Processing Systems: Natural and Synthetic. Cambridge: MIT Press, 2001: 849-856. |
24 | NIE F P, CAI G H, LI X L. Multi-view clustering and semi-supervised classification with adaptive neighbors[C]// Proceedings of the 31st AAAI Conference on Artificial Intelligence. Palo Alto, CA: AAAI Press, 2017: 2408-2414. 10.1609/aaai.v31i1.10909 |
25 | TAO Z Q, LIU H F, LI S, et al. From ensemble clustering to multi-view clustering[C]// Proceedings of the 26th International Joint Conference on Artificial Intelligence. California: ijcai.org, 2017:2843-2849. 10.24963/ijcai.2017/396 |
26 | MANNING C D, RAGHAVAN P, SCHÜTZE H. Introduction to Information Retrieval[M]. New York: Cambridge University Press, 2008: 100-103. |
27 | HONG M Y, LUO Z Q. On the linear convergence of the alternating direction method of multipliers[J]. Mathematical Programming, 2017, 162(1/2):165-199. 10.1007/s10107-016-1034-2 |
28 | LIN S X, ZHONG G, SHU T, et al. Simultaneously learning feature-wise weights and local structures for multi-view subspace clustering[J]. Knowledge-Based Systems, 2020, 205: No.106280. 10.1016/j.knosys.2020.106280 |
29 | KANG Z, ZHOU W T, ZHAO Z T. Large-scale multi-view subspace clustering in linear time[C]// Proceedings of the 34th AAAI Conference on Artificial Intelligence. Palo Alto, CA: AAAI Press, 2020: 4412-4419. 10.1609/aaai.v34i04.5867 |
30 | 辛利柯,杨琬琪,杨明. 基于判别稀疏性表示的不完整多视图分类[J]. 计算机科学与探索, 2021, 15(10):1938-1948. |
XIN L K, YANG W Q, YANG M. Incomplete multi-view classification via discriminative and sparse representation[J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(10):1938-1948. |
[1] | Shunyong LI, Shiyi LI, Rui XU, Xingwang ZHAO. Incomplete multi-view clustering algorithm based on self-attention fusion [J]. Journal of Computer Applications, 2024, 44(9): 2696-2703. |
[2] | Yao DONG, Yixue FU, Yongfeng DONG, Jin SHI, Chen CHEN. Survey of incomplete multi-view clustering [J]. Journal of Computer Applications, 2024, 44(6): 1673-1682. |
[3] | Chenghao YANG, Jie HU, Hongjun WANG, Bo PENG. Incomplete multi-view clustering algorithm based on attention mechanism [J]. Journal of Computer Applications, 2024, 44(12): 3784-3789. |
[4] | Yunhua ZHU, Bing KONG, Lihua ZHOU, Hongmei CHEN, Chongming BAO. Multi-view clustering network guided by graph contrastive learning [J]. Journal of Computer Applications, 2024, 44(10): 3267-3274. |
[5] | Ziyi HE, Yan YANG, Yiling ZHANG. Multi-view clustering network with deep fusion [J]. Journal of Computer Applications, 2023, 43(9): 2651-2656. |
[6] | Jinghuan LAO, Dong HUANG, Changdong WANG, Jianhuang LAI. Multi-view ensemble clustering algorithm based on view-wise mutual information weighting [J]. Journal of Computer Applications, 2023, 43(6): 1713-1718. |
[7] | LI Xingfeng, HUANG Yuqing, REN Zhenwen, LI Yihong. Robust multi-view clustering algorithm based on adaptive neighborhood [J]. Journal of Computer Applications, 2021, 41(4): 1093-1099. |
[8] | SONG Yan, YIN Jun. Multi-view spectral clustering algorithm based on shared nearest neighbor [J]. Journal of Computer Applications, 2020, 40(11): 3211-3216. |
[9] | CAO Dawei, HE Chaobo, CHEN Qimai, LIU Hai. Short text clustering algorithm based on weighted kernel nonnegative matrix factorization [J]. Journal of Computer Applications, 2018, 38(8): 2180-2184. |
[10] | QIU Baozhi, HE Yanfang, SHEN Xiangdong. Multi-view kernel K-means algorithm based on entropy weighting [J]. Journal of Computer Applications, 2016, 36(6): 1619-1623. |
[11] | DU Haijing XIAO Yanghui ZHU Dan TONG Xinxin. Affine-invariant shape matching algorithm based on modified multi-scale product Laplacian of Gaussian operator [J]. Journal of Computer Applications, 2014, 34(3): 841-845. |
Viewed | ||||||
Full text |
|
|||||
Abstract |
|
|||||