基于改进的混合学习模型的手写阿拉伯数字识别方法

徐琴珍; 杨绿溪

doi:10.3724/SP.J.1146.2009.00064

基于改进的混合学习模型的手写阿拉伯数字识别方法

doi: 10.3724/SP.J.1146.2009.00064

计量
- 文章访问数: 3542
- HTML全文浏览量: 86
- PDF下载量: 888
- 被引次数: 0
出版历程
- 收稿日期: 2009-01-06
- 修回日期: 2009-06-26
- 刊出日期: 2010-02-19

An Improved Hybrid Learning Model Based Handwritten Digits Recognition Approach

摘要

摘要: 在特征空间维数较高的手写阿拉伯数字识别问题中，冗余的特征往往会意外增加学习模型刻画问题空间的复杂度，影响手写阿拉伯数字识别的效率和精确度。该文提出了一种基于边界对特征的敏感度值进行特征选择的支持向量机树混合学习模型，依据当前中间节点上的分类曲面对子样本空间中的样例特征的敏感程度选择特征，在新构建的子样本集上训练子节点上的支持向量机。UCI机器学习数据库中手写阿拉伯数字识别问题的仿真结果表明，与其他算法相比，该文提出的方法能够在提高或保持手写阿拉伯数字高识别精确率的同时，精简问题空间，从而简化混合学习模型的中间节点和整体结构。
- 手写阿拉伯数字识别; 支持向量机树; 边界对特征的敏感度
Abstract: This paper is concerned with the problem of handwritten digits recognition in high dimensional feature space. The residual feature information may bring undesirable complexity to the underlying probability distribution of the concept label for learning algorithm to capture. The recognition accuracy and efficiency of the so trained learning model are usually depressed. According to this situation, an improved confusion-crossed support vector machine tree is proposed. A feature selection process based on the sensitivity of the margin to a feature is presented for the training step of each support vector machine embedded in each internal node. The experimental results on optical handwritten digits recognition problem in UCI database indicate that the proposed approach achieves competitive or even better recognition accuracy in the condensed feature space. Further, it also obtains lower structure complexity on internal nodes and the whole hybrid learning model than the compared approaches.

HTML全文

参考文献(1)

Awaidah S M and Mahmoud S A. A multiple feature/resolution scheme to Arabic (Indian) numerals recognition using hidden Markov models[J].Signal Processing.2009, 89(6):1176-1184[2]Miao Kang and Dominic Palmer-Brown. A modal learning adaptive function neural network applied to handwritten digit recognition[J].Information Sciences.2008, 178(20):3802-3812[3]Polat ? and Yildirim T. Genetic optimization of GRNN for pattern recognition without feature extraction[J].Expert Systems with Applications.2008, 34(4):2444-2448[4]Sabri Mahmoud. Recognition of writer-independent off-line handwritten Arabic (Indian) numerals using hidden Markov models[J].Signal Processing.2008, 88(4):844-857[5]徐琴珍, 章品正, 裴文江, 杨绿溪, 何振亚. 基于混淆交叉支撑向量机树的自动面部表情分类方法. 中国图象图形学报, 2008, 13(7): 1329-1334.Xu Qin-zhen, Zhang Pin-zheng, Pei Wen-jiang, Yang Lu-xi, and He Zhen-ya. An automatic facial expression recognition approach based on confusion-crossed support vector machine tree. Journal of Image and Graphics, 2008, 13(7): 1329-1334.[6]Melgani F and Bazi Y. Classification of electrocardiogram signals with support vector machines and particle swarm optimization[J].IEEE Transactions on Information Technology in Biomedicine.2008, 12(5):667-677[7]Sindhwani V, Rakshit S, Deodhare D, Erdogmus D, Principe J C, and Nivogi P. Feature selection in MLPs and SVMs based on maximum output information[J].IEEE Transactions on Neural Networks.2004, 15(4):937-948[8]Bo L F, Wang L, and Jiao L C. Training hard-margin support vector machines using greedy stagewise algorithm[J].IEEE Transactions on Neural Networks.2008, 19(8):1446-1455[9]徐琴珍. 树型混合学习模型及其应用研究. [博士论文], 东南大学, 2006.[10]Xu Qin-zhen. Approach and application research of tree-structured hybrid learning models. [Ph. D. dissertation], Southeast University, 2006.[11]Saul L K and Roweis S T. Think globally, fit locally: Unsupervised learning of low dimensional manifolds[J].Journal of Machine Learning Research.2004, 4(2):119-155[12]Quilan J R. C4.5: Programs for Machine Learning. San Mateo: Morgan Kaufmann Publishers, 1993: 17-26.[13]Bennett K P, Shawe-Taylor J, and Demiriz A. Linear programming boosting via column generation[J].Machine Learning.2001, 46(1):225-254[14]Hsu C W and Lin C J. A comparison of methods for multiclass support vector machines[J].IEEE Transactions on Neural Networks.2002, 13(2):415-525

施引文献

资源附件(0)

访问统计