Full Text Views
18
Many kinds of data can be represented as a network or graph. It is crucial to infer the latent structure underlying such a network and to predict unobserved links in the network. Mixed Membership Stochastic Blockmodel (MMSB) is a promising model for network data. Latent variables and unknown parameters in MMSB have been estimated through Bayesian inference with the entire network; however, it is important to estimate them online for evolving networks. In this paper, we first develop online inference methods for MMSB through sequential Monte Carlo methods, also known as particle filters. We then extend them for time-evolving networks, taking into account the temporal dependency of the network structure. We demonstrate through experiments that the time-dependent particle filter outperformed several baselines in terms of prediction performance in an online condition.
Tomoki KOBAYASHI
Kobe University
Koji EGUCHI
Kobe University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Tomoki KOBAYASHI, Koji EGUCHI, "Online Inference of Mixed Membership Stochastic Blockmodels for Network Data Streams" in IEICE TRANSACTIONS on Information,
vol. E97-D, no. 4, pp. 752-761, April 2014, doi: 10.1587/transinf.E97.D.752.
Abstract: Many kinds of data can be represented as a network or graph. It is crucial to infer the latent structure underlying such a network and to predict unobserved links in the network. Mixed Membership Stochastic Blockmodel (MMSB) is a promising model for network data. Latent variables and unknown parameters in MMSB have been estimated through Bayesian inference with the entire network; however, it is important to estimate them online for evolving networks. In this paper, we first develop online inference methods for MMSB through sequential Monte Carlo methods, also known as particle filters. We then extend them for time-evolving networks, taking into account the temporal dependency of the network structure. We demonstrate through experiments that the time-dependent particle filter outperformed several baselines in terms of prediction performance in an online condition.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E97.D.752/_p
Copy
@ARTICLE{e97-d_4_752,
author={Tomoki KOBAYASHI, Koji EGUCHI, },
journal={IEICE TRANSACTIONS on Information},
title={Online Inference of Mixed Membership Stochastic Blockmodels for Network Data Streams},
year={2014},
volume={E97-D},
number={4},
pages={752-761},
abstract={Many kinds of data can be represented as a network or graph. It is crucial to infer the latent structure underlying such a network and to predict unobserved links in the network. Mixed Membership Stochastic Blockmodel (MMSB) is a promising model for network data. Latent variables and unknown parameters in MMSB have been estimated through Bayesian inference with the entire network; however, it is important to estimate them online for evolving networks. In this paper, we first develop online inference methods for MMSB through sequential Monte Carlo methods, also known as particle filters. We then extend them for time-evolving networks, taking into account the temporal dependency of the network structure. We demonstrate through experiments that the time-dependent particle filter outperformed several baselines in terms of prediction performance in an online condition.},
keywords={},
doi={10.1587/transinf.E97.D.752},
ISSN={1745-1361},
month={April},}
Copy
TY - JOUR
TI - Online Inference of Mixed Membership Stochastic Blockmodels for Network Data Streams
T2 - IEICE TRANSACTIONS on Information
SP - 752
EP - 761
AU - Tomoki KOBAYASHI
AU - Koji EGUCHI
PY - 2014
DO - 10.1587/transinf.E97.D.752
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E97-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2014
AB - Many kinds of data can be represented as a network or graph. It is crucial to infer the latent structure underlying such a network and to predict unobserved links in the network. Mixed Membership Stochastic Blockmodel (MMSB) is a promising model for network data. Latent variables and unknown parameters in MMSB have been estimated through Bayesian inference with the entire network; however, it is important to estimate them online for evolving networks. In this paper, we first develop online inference methods for MMSB through sequential Monte Carlo methods, also known as particle filters. We then extend them for time-evolving networks, taking into account the temporal dependency of the network structure. We demonstrate through experiments that the time-dependent particle filter outperformed several baselines in terms of prediction performance in an online condition.
ER -