南京大学学报(自然科学版) ›› 2012, Vol. 48 ›› Issue (2): 133139.
解晓敏1,李云2**
Xie Xiao-Min1 ,Li Yun2
摘要: 利用最小最大模块化网络实现模式分类的关键问题之一就是找到一种有效且复杂度较低的训练样木划分方法,以便缩短训练的时间,得到相对平衡的划分子集.木文提出一种新的基于二分K-均值的训练集划分方法,它可以得到全局最优解,时间复杂度较低,并且可以通过层次聚类得到相对平衡的样木划分效果.在现实数据集上的实验表明,该划分方法在不降低分类精确率的情况卜能有效地缩短最小最大模块化网络的训练时间.
[1]Provost F, Aronis J M. Scaling up inductive learning with massive parallelism. Machine Learning, 1996,23:33一46. [2]Lu B L,Shin J H,lchikawa M. Massively par- allel classification of single-trial EEG signals using a Min-Max modular neural network. IEEE Transactions on Biomedical Engineering, 2004,51(3):551一558. [3]Ma Q, Lu B L,lsahara H,et al. Part oL speech tagging with Min-Max modular neural networks. Systems and Computers in Japan, 2002,33(7):30一39. [4]Shin J H,Lu Y I.,Talnov A,et al. Reading auditory discrimination behavior of freely mov- ing rats from hippocampal EEG. Neurocomput ing, 2001,(38):1557一1566. [5]Ma Q, Lu B L, Murata M, et al. Online error detection of annotated corpus using modular neural networks. Proceedings of Artificial neu- ral networks, Lecture Notes in Computer Scince, Springcr-Verlag, 2001,2130: 118一1192. [6]Huang B, Lu B L. Fault diagnosis for industrial images using a Min-Max modular neural net- work. Proceedings of Neural information Pro- cessing, Lecture Notes in Computer Scince,Springer-Verlag, 2004,3316;842一847. [7]Liu F Y,Wu K,Zhao H,et al. Fast text cater gorization with Min-Max modular support vec for machines. Proceedings of the international Joint Conference on Neural Networks, 2005: 570一575. [8]Fan Z G, Lu B L. Multi-view face recognition with Min-Max modular SVMs. Proceedings of Advances in Natural Computation;The 1st1n- ternational Conference, Lecture Notes in Com- puter Scince, Springcr-Verlag, 2005,3611: 396一399. [9]Lian H C,Lu B L. Multi-view gender classifi- canon using local binary patterns and support vector machines. Proceedings of the 3rdlnterna- tional Symposium on Neural Networks, Lecture Notes in Computer Scince, Springer-Verlag, 2006,3972:202一209. [10]Yang Y,Lu B L. Prediction of protein sub cel lular multi-locations with a Min-Max modular support vector machine. Proceedings of the 3rd international Symposium on Neural Networks,Lecture Notes in Computer Scince, Springer- Verlag, 2006,3973:667一673. [11]Wang K A. Min-Max modular support vector machine and its application to text classification. Master Thesis. Shanghai:Shanghai Jiao Tong University, 2005.(土开安.最小最大模块化支持向量机及其在文木分类中的应用.硕士学位论文.上海:上海交通大学,2005). [12]Wen Y M, Lu B L, Zhao H. Equal clustering makes Min-Max modular support vector ma- chine more efficient. Proceedings of Intcrnation- al Conference on Neural information Process- ing, Taipei,2005,77一82. [13]Luo J. Face image based gender classification u- sing Min-Max modular classifier. Master The- sis. Shanghai:Shanghai Jiao Tong University, 2007.(罗俊.基于最小最大模块化分类器的人脸图像的性别分类研究.硕士学位论文.上海:上海交通大学,2007). [14]Lian H C, Lu B L,Takikawa E, et al. Gender recognition using a Min-Max modular support vector machine. Proceedings of the 10’ lnterna- tional Conference on the Advances in Natural Computation, Lecture Notes in Computer Scincc, Springcr-Vcrlag, 2005,3611: 438一44l. [15]Lu B L, M, lto.Task decomposition and mod- ule combination based on class relations; A modular neural network for pattern classifica- tion. IEEE Transactions on Neural Networks,1999,10(5):1224~1256 [16]Zheng M M, Ji G L. An improved density based distributed clustering. Journal of Nanjing University(Natural Sciences),2008,44(5): 536-543.(郑苗苗,吉根林.一种基于密度的分布式聚类算法.南京大学学报(自然科学),2008,44(5):536一543). [17]Han J,Kamber M. Data mining; Concepts and techniques. Morgan Kaufmann. Fan M, Meng X F. Beijing; China Machine Press,2001,374. (Han J , Kambcr M.数据挖掘:概念与技术.范明,孟小峰.北京:机械工业出版社,2001,3 74). [18]Stcinbach M,Karypis G, Kumar V. A Com- parison of document clustering techniques. Proceedings of Conference Data Mining and Knowledge Discovery, Workshop on Tcxt Min- ing, Boston, 2000,1一20. [19]Zhao Y,KaryPis G. Hierarchical clustering al- gorithms for document datasets. Proceedings of Conference Data Mining and Knowledge Discov- ery, Data Mining and Knowledge Biscouevy, Springer-Vcrlag, 2005,10(2):141一168. [20]Du C H,Huang X Y,Yang Z Y,et al. Appli- canon of inproved fuzzy C-means clustering in automatic programming traffic intervals. Com- puter Engineering and Applications, 2009,45 (24) ; 190- 193.(杜长海,黄席抛,杨祖元等. 改进的FCM聚类在交通时段自动划分中的应 用.计算机工程与应用,2009,45(24); 190~ 193) |
No related articles found! |
|