Multi-HDP: A Non Parametric Bayesian Model for Tensor Factorization
Multi-HDP: A Non Parametric Bayesian Model for Tensor Factorization
复制标题
DOI：
--
复制
发表时间：
2008-07
期刊：
影响因子：
--
通讯作者：
I. Porteous;E. Bart;M. Welling
中科院分区：
文献类型：
--
作者：
I. Porteous;E. Bart;M. Welling
矩阵分解算法在机器学习社区中经常被用来寻找数据的低维表示。我们介绍了一种新的生成贝叶斯概率模型的无监督矩阵和张量分解。该模型由多个相互作用的LDA模型组成，每个模型对应一个模态。我们描述了一个有效的折叠吉布斯采样推理。我们还推导出非参数形式的模型，其中相互作用的LDA模型被替换为相互作用的HDP模型。实验表明，该模型是有用的预测缺失的数据与两个或两个以上的模态，以及学习数据中的潜在结构。
Matrix factorization algorithms are frequently used in the machine leaming community to find low dimensional representations of data. We introduce a novel generative Bayesian probabilistic model for unsupervised matrix and tensor factorization. The model consists of several interacting LDA models, one for each modality. We describe an efficient collapsed Gibbs sampler for inference. We also derive the non-parametric form of the model where interacting LDA models are replaced with interacting HDP models. Experiments demonstrate that the model is useful for prediction of missing data with two or more modalities as well as learning the latent structure in the data.