Abstract—Bayesian model of inference is widely used in
various application fields such as data engineering or text
processing. Using Bayes’ theorem, we can obtain the posterior
distribution function. When we conduct sampling using
Markov chain Monte Carlo (MCMC), the most prominent
MCMC algorithms are the Metropolis-Hastings and the Gibbs
sampler, the latter being particularly useful in Bayesian
analysis. This paper presents the visual teaching material for
studying Gibbs sampler algorithm. Interaction with this
material is supposed to enable students to deeply understand
the mathematical process behind Gibbs sampling and
encourages them to comprehend the mathematical expressions
in the textbooks.
Index Terms—Bayesian inference, MCMC, simple topic
model, Gibbs sampler, visualization.
Y. Shirota is with Faculty of Economics, GakushuinUniversityy, 1-5-1
Mejiro, Toshima-ku Tokyo, 171-8588 Japan (e-mail:
yukari.shirota-atmark-gakushuin.ac.jp).
T. Hashimoto is with Chiba University of Commerce, 1-3-1, Konodai
Ichikawa City, Chiba, 272-8512, Japan (e-mail: takako-atmark-cuc.ac.jp).
B. Chakraborty is with the Department of Software and Information
Science, Iwate Prefectural University, 152-52 Sugo, Takizawa, Iwate
020-0693, Japan (e-mail: basabi-atmark-iwate-pu.ac.jp).
[PDF]
Cite: Yukari Shirota, Takako Hashimoto, and Basabi Chakraborty, "Visual Materials to Teach Gibbs Sampler," International Journal of Knowledge Engineering vol. 2, no. 2, pp. 92-95, 2016.
