We develop a class of random effect models for dynamic networks aimed to account for a time-varying network topology. Ignoring this aspect may lead to a distortion in the parameter estimate of the model for the edges, since, in principle, an edge is not observable when the related couple of nodes does not belong to the network. A Bayesian conjugate approach is proposed for the inference of this class of models based on the P´olya-Gamma latent variable method. A preliminary simulation study has
Dynamic network models with time-varying nodes / Gherardini L., Bernardi M., Lupparelli M.. - ELETTRONICO. - (2023), pp. 1-6. (Intervento presentato al convegno Statistical Learning, Sustainability and Impact Evaluation).
Dynamic network models with time-varying nodes
Lupparelli M
2023
Abstract
We develop a class of random effect models for dynamic networks aimed to account for a time-varying network topology. Ignoring this aspect may lead to a distortion in the parameter estimate of the model for the edges, since, in principle, an edge is not observable when the related couple of nodes does not belong to the network. A Bayesian conjugate approach is proposed for the inference of this class of models based on the P´olya-Gamma latent variable method. A preliminary simulation study has
| File | Dimensione | Formato | |
|---|---|---|---|
|
LucaGherardini_SIS2023.pdf
Accesso chiuso
Tipologia:
Pdf editoriale (Version of record)
Licenza:
Tutti i diritti riservati
Dimensione
223.9 kB
Formato
Adobe PDF
|
223.9 kB | Adobe PDF | Richiedi una copia |
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
