Titlebook: Probabilistic Topic Models; Foundation and Appli Di Jiang,Chen Zhang,Yuanfeng Song Book 2023 The Editor(s) (if applicable) and The Author(s

不能仁慈

CRAFT

Topic Models, to a wide range of tasks. In this chapter, we select some representative topic models for introducing the mathematical principles behind topic models. By studying this chapter, readers will have a deep understanding of the foundation of topic models, and the ability to select appropriate existing m

NICHE

Expectation Maximization,hod used to obtain the local optima of these parameters of probabilistic graphical models with latent variables. This chapter introduces how to apply the EM algorithm in topic models such as PLSA. The foundation of EM algorithm is introduced in Sect. 4.1. The convergence of EM is introduced in Sect.

浪蕩子

anchor

apiary

Distributed Training, to train such large-scale topic models encounters bottlenecks in computing efficiency and data storage. Therefore, it is necessary to develop distributed training mechanisms for topic models. In this chapter, we introduce distributed computing architectures in Sect. 7.1, followed by the distributed

觀察

Topic Models,. By studying this chapter, readers will have a deep understanding of the foundation of topic models, and the ability to select appropriate existing models or design brand-new models for their own scenarios.

ARK

Expectation Maximization,the EM algorithm in topic models such as PLSA. The foundation of EM algorithm is introduced in Sect. 4.1. The convergence of EM is introduced in Sect. 4.2 and the generalized expectation maximization (GEM) is discussed in Sect. 4.3. Finally, the applications of EM and GEM in topic models are explained in Sect. 4.4.

Asperity

概觀