首页|Source estimation in continuous-time diffusion networks via incomplete observation

Source estimation in continuous-time diffusion networks via incomplete observation

扫码查看
原文链接
  • NSTL
  • Elsevier
This paper considers the problem of estimating the source of diffusion in a network under incomplete observation condition. The diffusion process is described by a continuous-time information diffusion model and the source estimation is formulated as a maximum likelihood (ML) estimator in terms of a Gaussian weighted averaging of the correlation coefficients between the observed activation times and the sampled transmission delays obtained by Monte Carlo simulations. Experiments are worked out with both synthetic and real-world networks to show the effectiveness of our method in comparison with previous results.& nbsp;(C)& nbsp;& nbsp;2021 Elsevier B.V. All rights reserved.

Diffusive networkSource estimationTransmission delay distributionGaussian weighted averaging correlation & nbspcoefficientMonte Carlo simulation

Shi, Chaoyi、Zhang, Qi、Chu, Tianguang

展开 >

Peking Univ

Univ Int Business & Econ

2022

10.1016/j.physa.2021.126843

Physica

Physica

ISSN：0378-4371
年,卷(期)：2022.592
  • 4
  • 34