广义狄利克雷型分布不完全分类数据的统计分析及贝叶斯抽样

在社会学、心理学、保险学和流行病学等学科中,研究人员经常利用类别变量来研究现象的分布特征或者因素之间的关联。本文提出在类别变量的研究中存在一类不完全分类数据,其观测数据的似然函数具有广义狄利克雷型分布的形式。首先,利用EM算法和一种新的基于共众数思想的优化方法计算参数的极大似然估计。其次,在贝叶斯分析中,建立新的结合等高共众数方法的采样重要性重抽样算法来实现广义狄利克雷型分布的有效后验样本抽样。并将所提出的方法用于两组实例数据的分析,实证分析结果表明了本文所提出的方法在一般的不完全分类数据分析中的有效性和实用性。

狄氏型及其在数学物理中的应用

简介狄氏型理论。该理论已成为紧密联系解析位势论与马氏过程理论的强有力数学工具,并因此在数学与物理中有许多应用。

Semantic Knowledge Acquisition from Blogs with Tag-Topic Model

This paper focuses on semantic knowl- edge acquisition from blogs with the proposed tag- topic model. The model extends the Latent Dirichlet Allocation （LDA） model by adding a tag layer be- tween the document and the topic. Each document is represented by a mixture of tags; each tag is as- sociated with a multinomial distribution over topics and each topic is associated with a multinomial dis- trNution over words. After parameter estimation, the tags are used to descrNe the underlying topics. Thus the latent semantic knowledge within the top- ics could be represented explicitly. The tags are treated as concepts, and the top-N words from the top topics are selected as related words of the con- cepts. Then PMI-IR is employed to compute the re- latedness between each tag-word pair and noisy words with low correlation removed to improve the quality of the semantic knowledge. Experiment re- sults show that the proposed method can effectively capture semantic knowledge, especially the polyse- me and synonym.

Mining User Interest in Microblogs with a User-Topic Model

Microblogs have become an important platform for people to publish,transform information and acquire knowledge.This paper focuses on the problem of discovering user interest in microblogs.In this paper,we propose a topic mining model based on Latent Dirichlet Allocation(LDA) named user-topic model.For each user,the interests are divided into two parts by different ways to generate the microblogs:original interest and retweet interest.We represent a Gibbs sampling implementation for inference the parameters of our model,and discover not only user's original interest,but also retweet interest.Then we combine original interest and retweet interest to compute interest words for users.Experiments on a dataset of Sina microblogs demonstrate that our model is able to discover user interest effectively and outperforms existing topic models in this task.And we find that original interest and retweet interest are similar and the topics of interest contain user labels.The interest words discovered by our model reflect user labels,but range is much broader.

Self-Adaptive Topic Model： A Solution to the Problem of ＂Rich Topics Get Richer＂

The problem of "rich topics get richer"(RTGR) is popular to the topic models,which will bring the wrong topic distribution if the distributing process has not been intervened.In standard LDA(Latent Dirichlet Allocation) model,each word in all the documents has the same statistical ability.In fact,the words have different impact towards different topics.Under the guidance of this thought,we extend ILDA(Infinite LDA) by considering the bias role of words to divide the topics.We propose a self-adaptive topic model to overcome the RTGR problem specifically.The model proposed in this paper is adapted to three questions:(1) the topic number is changeable with the collection of the documents,which is suitable for the dynamic data;(2) the words have discriminating attributes to topic distribution;(3) a selfadaptive method is used to realize the automatic re-sampling.To verify our model,we design a topic evolution analysis system which can realize the following functions:the topic classification in each cycle,the topic correlation in the adjacent cycles and the strength calculation of the sub topics in the order.The experiment both on NIPS corpus and our self-built news collections showed that the system could meet the given demand,the result was feasible.

Dirichlet boundary value problem with variable growth

In this paper, we study higher order elliptic partial differential equations with variable growth, and obtain the existence of solutions in the setting of Wm,p（x） spaces by means of an abstract result for variational inequalities obtained by Gossez and Mustonen. Our result generalizes the corresponding one of Kováik and Rákosník.

Topic modeling for large-scale text data

This paper develops a novel online algorithm, namely moving average stochastic variational inference （MASVI）, which applies the results obtained by previous iterations to smooth out noisy natural gradients. We analyze the convergence property of the proposed algorithm and conduct a set of experiments on two large-scale collections that contain millions of documents. Experimental results indicate that in contrast to algorithms named ＇stochastic variational inference＇ and ＇SGRLD＇, our algorithm achieves a faster convergence rate and better performance.