摘要
针对人体行为识别过程中分类算法识别精度低和数据样本集的"维数灾难"问题,提出了基于行为识别的SPLDA降维算法.首先,利用SPLDA算法在原有样本协方差矩阵不变的情况下获取最重要的主分量,通过贪婪搜索方法得到多个投影向量;然后,通过更新类内散度矩阵获得最优转换矩阵;最后,将降维后的样本数据集通过XGBoost分类器进行最终的行为识别.实验结果表明,XGBoost分类器与随机森林算法相比,平均识别精度提高了2.66%,识别时间降低了0.52 s;SPLDA-XGB算法可以实现有效降维且比PCA算法、LDA算法、LPP算法、L-PCA算法与XGBoost分类器结合的识别算法具有更高的人体行为识别准确率.
Aiming at the problem of "dimension disaster" in human behavior recognition and the classification algorithm has low recognition accuracy and data sample set. First, the SPLDA algorithm is used to obtain the most important principal components with the original sample covariance matrix unchanged, and multiple projection vectors are obtained by greedy search method. Then, the optimal transformation matrix is obtained by updating the class inner divergence matrix. Finally, the dimensionally reduced sample data set is identified by the XGBoost classifier. Experimental results show that compared with the random forest algorithm, the average recognition accuracy of XGBoost classifier is improved by 2.66% and the recognition time is reduced by 0.52 s. SPLDA-XGB algorithm can achieve effective dimensionality reduction and has higher accuracy rate of human behavior recognition than PCA algorithm, LDA algorithm, LPP algorithm, l-pca algorithm combined with XGBoost classifier.
引文
[1] KHAN A M,TUFAIL A,KHATTAK A M.Ativity recognition on smartphones via sensor-fusion and KDA-based SVMs[J].International Journal of Distributed Sensor Networks,2014(1):1-14.
[2] KWAPIAZ J R,WEISS G M,MOORE S A.Activ-ity recognition using cell phoneaccele ro-meters[J].ACM SigKDD Explorations Newsle-tter,2011,12(2):74-82.
[3] YANG J,WRIGHT J,HUANG T S,et al.Image s-uperresolution via sparse representation[J].IEEE Transactions on Image ProcessingA Publication of the IEEE Signal Pocessing So-ciety,2010,19(11):28-128.
[4] 聂军.基于K-L特征压缩的云计算冗余数据降维算法[J].微电子学与计算机,2016,33(2):125-129.
[5] VAN D M L,POSTMA E,VAN DENIIERIK J.Dimension reduction:A comparative review[R].TiCCTR:Tiburg University,2009.
[6] 李龙龙,何东健,王美丽.L-PCA算法下的高维图像降维算法研究[J].西安科技大学学报,2017,37(6):906-911.
[7] 钟福金.鲁棒线性子空间学习算法与框架研究[D]:成都:西南交通大学,2015.
[8] Chen T Q.XGBoost:A Scalable Tree Boosting System[C].In Proceeding KDD'16 Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.San Francisco,2016:2-4.
[9] Wang H X,Qin T,Zheng W M.L1-norm-based common spatial patterns.[J].IEEE Transactions on Biomedical Engineering,2012,59(3):653-662.