Algorithms of the LDA model [REPORT]
This is an incremental review for researchers working with LDA models, focusing on algorithm efficiency.
The paper reviews three algorithms for Latent Dirichlet Allocation (LDA), comparing their time complexity and performance, and finds that online variational Bayesian inference is the fastest while still delivering reasonably good results.
We review three algorithms for Latent Dirichlet Allocation (LDA). Two of them are variational inference algorithms: Variational Bayesian inference and Online Variational Bayesian inference and one is Markov Chain Monte Carlo (MCMC) algorithm -- Collapsed Gibbs sampling. We compare their time complexity and performance. We find that online variational Bayesian inference is the fastest algorithm and still returns reasonably good results.