Adversarial Learning for Topic Models
| Autor: | Tomonari Masada, Atsuhiro Takasu |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Topic model
Perplexity Discriminator Artificial neural network Computer science business.industry Posterior probability Probabilistic logic 02 engineering and technology 010501 environmental sciences Machine learning computer.software_genre 01 natural sciences Latent Dirichlet allocation symbols.namesake ComputingMethodologies_PATTERNRECOGNITION 020204 information systems Prior probability 0202 electrical engineering electronic engineering information engineering symbols Artificial intelligence business computer 0105 earth and related environmental sciences |
| Zdroj: | Advanced Data Mining and Applications ISBN: 9783030050894 ADMA |
| Popis: | This paper proposes adversarial learning for topic models. Adversarial learning we consider here is a method of density ratio estimation using a neural network called discriminator. In generative adversarial networks (GANs) we train discriminator for estimating the density ratio between the true data distribution and the generator distribution. Also in variational inference (VI) for Bayesian probabilistic models we can train discriminator for estimating the density ratio between the approximate posterior distribution and the prior distribution. With the adversarial learning in VI we can adopt implicit distribution as an approximate posterior. This paper proposes adversarial learning for latent Dirichlet allocation (LDA) to improve the expressiveness of the approximate posterior. Our experimental results showed that the quality of extracted topics was improved in terms of test perplexity. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|