Infinite latent feature models and the Indian buffet process
Infinite latent feature models and the Indian buffet process
复制标题
DOI：
--
复制
发表时间：
2005-12
期刊：
影响因子：
--
通讯作者：
T. Griffiths;Zoubin Ghahramani
中科院分区：
文献类型：
--
作者：
T. Griffiths;Zoubin Ghahramani
我们定义了一个概率分布在等价类的二元矩阵有限的行数和无限的列数。该分布适合用作概率模型中的先验，该概率模型使用潜在的无限特征阵列来表示对象。我们确定了一个简单的生成过程，结果在相同的分布在等价类，我们称之为印度自助餐过程。我们说明了使用这种分布作为一个前在无限的潜在特征模型，推导出马尔可夫链蒙特卡罗算法在这个模型中的推理和应用该算法的图像数据集。
We define a probability distribution over equivalence classes of binary matrices with a finite number of rows and an unbounded number of columns. This distribution is suitable for use as a prior in probabilistic models that represent objects using a potentially infinite array of features. We identify a simple generative process that results in the same distribution over equivalence classes, which we call the Indian buffet process. We illustrate the use of this distribution as a prior in an infinite latent feature model, deriving a Markov chain Monte Carlo algorithm for inference in this model and applying the algorithm to an image dataset.