期刊文献+

HDP消息传递算法

Belief Propagation Algorithm for HDP
下载PDF
导出
摘要 分层狄利克雷过程(HDP)可以视为潜在狄利克雷分配(LDA)主题模型在无参方面的扩展,解决了传统主题模型中选择主题数目的问题.消息传递算法则是从消息因子图模型的角度出发进而解决贝叶斯后验概率推断问题.提出将消息传递算法应用到分层狄利克雷过程模型上的方案,并从最大期望算法的角度来证明该算法的收敛性,最终从混淆度的角度将该算法与传统算法进行对比.实验结果表明消息传递算法在混淆度方面相比其他算法有明显的优势,且收敛速度较快. The hierarchical Dirichlet process (HDP) model is an extension of the latent Dirichlet allocation(LDA)on the aspect of non-parametric in order to solve the problem of setting number of the topics. Belief propagation algorithm is an algorithm based on the factor graph model to inference the Bayesian posterior probability. In our paper, we propose to apply the belief propagation algorithm on the HDP model, and prove the convergence of the algorithm from the view of expectation maximization algorithm. Comparing with other algorithms, the belief propagation algorithm based on HDP is better than others in accuracy measured by perplexity
出处 《微电子学与计算机》 CSCD 北大核心 2016年第3期142-146,151,共6页 Microelectronics & Computer
基金 国家自然科学基金(61373092 61033013 61272449 61202029) 江苏省教育厅重大项目(12KJA520004) 江苏省科技支撑计划重点项目(BE2014005) 广东省重点实验室开放课题(SZU-GDPHPCL-2012-09)
关键词 分层狄利克雷过程 消息传递算法 无参数主题模型 最大期望算法 shierarchical Dirichlet process belief propagation algorithm non-parametric topic model expectation maximize
  • 相关文献

参考文献7

  • 1BleiD M, Ng A Y, Jordan M 1. I.atent dirichlet atloca tion[J]. The Journal of Uachine Learning Research, 2003,3(1): 993-1022. 被引量:1
  • 2Gershman S J, Blei D M. A tutorial on Bayesian non- parametric models[J]. Journal of Mathematical Psy ehology, 2012, 56(1): 1-12. 被引量:1
  • 3Zeng J, Cheung W K, Liu J. Learning topic models by belief propagation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35 ( 5 ) 1121-1134. 被引量:1
  • 4周建英,王飞跃,曾大军.分层Dirichlet过程及其应用综述[J].自动化学报,2011,37(4):389-407. 被引量:40
  • 5Paisley J W, Zaas A K, Woods C W, et al. A stick- breaking construction of the beta process [C] ff Pro- ceedings of the 27th International Conference on Ma- chine Learning. Haifa, Isreal.. IEEE, 2010. 被引量:1
  • 6Teh Y W, Kurihara K, Welling M. Collapsed variational inference for HDP[C] /// Advances in neural in- formation processing systems. Is. 1. ]:IEEE,2007. 被引量:1
  • 7HannaA M, Wallach M, David M Mimmo. Rethin- king lda: Why priors matter[C]// Annual Conference on Neural Information Processing Systems. Vancou-ver: Curran Associates, 2009:1973-1981. 被引量:1

二级参考文献98

  • 1Mitchell T M. Machine Learning. New York: McGraw-Hill, 1997. 被引量:1
  • 2Teh Y W. Dirichlet processes. Encyclopedia of Machine Learning, Springer, 2010. Part 5, 280-287. 被引量:1
  • 3Teh Y W, Jordan M I. Hierarchical Bayesian nonparametric models with applications. Bayesian Nonparametrics Princi- ples and Practice. Cambridge University Press, 2009. 1-47. 被引量:1
  • 4Teh Y W, Jordan M I, Beal M J, Blei D M. Sharing clus- ters among related groups: hierarchical Dirichlet processes. In: Proceedings of the Advances in Neural Information Processing Systems. Vancouver, Canada: The MIT Press, 2004. 1385 - 1392. 被引量:1
  • 5Teh Y W, Jordan M I, Beal M J, Blei D M. Hierarchical Dirichlet processes. Journal of the American Statistical As- sociation, 2006, 101(476): 1566-1581. 被引量:1
  • 6Yakhnenko O, Honavar V. Multi-modal hierarchical Dirich- let process model for predicting image annotation and image-object label correspondence. In: Proceedings of the SIAM International Conference on Data Mining. Sparks, USA: SIAM, 2009. 281-294. 被引量:1
  • 7Wang X G, Ma X K, Grimson W E L. Unsupervised activity perception by hierarchical Bayesian models. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Minneapolis, USA: IEEE, 2007. 1-8. 被引量:1
  • 8Wang X, Tieu K, Gee-Wah N, Grimson W E L. Trajectory analysis and semantic region modeling using a nonpaxamet- ric Bayesian model. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA: IEEE, 2008. 1-8. 被引量:1
  • 9Wang X G, Ma X X, Grimson W E L. Unsupervised activity perception in crowded and complicated scenes using hierarchical Bayesian models. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(3): 539-555. 被引量:1
  • 10Wang X G, Grimson W E L, Westin C F. Tractography segmentation using a hierarchical Dirichlet processes mixture model. In: Proceedings of the 21st International Conference on Information Processing in Medical Imaging. Williams- burg, USA: Springer, 2009. 101-113. 被引量:1

共引文献39

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部