Modeling and inferring Sampling design in probabilistic random network models (2016-2019)
Timothée’s PhD was co-supervised with Pierre Barbillon (50%/50%), from 2016 to 2019.
Summary
This thesis studies the Stochastic Block Model (SBM) for network data collected under an imperfect, partially observed sampling design. Building on a variational inference framework, it develops estimators that explicitly account for the sampling process, so that latent block structure can be recovered without inferring artefactual clusters induced by missing pairs of observations. The thesis introduces a generic characterization of sampling designs and studies their identifiability with respect to the SBM parameters, and provides the missSBM R package implementing the resulting methodology. Applications are given to seed exchange networks and ecological interaction networks, where sampling is typically incomplete and non-random.
PhD manucript
2019
PhD thesis
Impact de l’échantillonnage sur l’inférence de structures dans les réseaux: application aux réseaux d’échanges de graines et à l’écologie
This thesis studies the Stochastic Block Model (SBM) for network data collected under an imperfect, partially observed sampling design. Building on a variational inference framework, it develops estimators that explicitly account for the sampling process, so that latent block structure can be recovered without inferring artefactual clusters induced by missing pairs of observations. The thesis introduces a generic characterization of sampling designs and studies their identifiability with respect to the SBM parameters, and provides the missSBM R package implementing the resulting methodology. Applications are given to seed exchange networks and ecological interaction networks, where sampling is typically incomplete and non-random.
Journal papers
2022
J. Stat. Softw.
missSBM: An R Package for Handling Missing Values in the Stochastic Block Model
Pierre Barbillon, Julien Chiquet, and Timothée Tabouy
<p>The stochastic block model is a popular probabilistic model for random graphs. It is commonly used to cluster network data by aggregating nodes that share similar connectivity patterns into blocks. When fitting a stochastic block model to a partially observed network, it is important to consider the underlying process that generates the missing values, otherwise the inference may be biased. This paper presents missSBM, an R package that fits stochastic block models when the network is partially observed, i.e., the adjacency matrix contains not only 1s or 0s encoding the presence or absence of edges, but also NAs encoding the missing information between pairs of nodes. This package implements a set of algorithms to adjust the binary stochastic block model, possibly in the presence of external covariates, by performing variational inference suitable for several observation processes. Our implementation automatically explores different block numbers to select the most relevant model according to the integrated classification likelihood criterion. The integrated classification likelihood criterion can also help determine which observation process best fits a given dataset. Finally, missSBM can be used to perform imputation of missing entries in the adjacency matrix. We illustrate the package on a network dataset consisting of interactions between political blogs sampled during the 2007 French presidential election.</p>
@article{JSSv101i12,title={missSBM: An R Package for Handling Missing Values in
the Stochastic Block Model},volume={101},doi={10.18637/jss.v101.i12},number={12},journal={Journal of Statistical Software},author={Barbillon, Pierre and Chiquet, Julien and Tabouy, Timothée},year={2022},pages={1–32},}
2020
JASA
Variational Inference for Stochastic Block Models from Sampled Data
Timothée Tabouy, Pierre Barbillon, and Julien Chiquet
Journal of the American Statistical Association, 2020
@article{2019_JASA_TBC,author={Tabouy, Timothée and Barbillon, Pierre and Chiquet, Julien},title={Variational Inference for Stochastic Block Models
from Sampled Data},journal={Journal of the American Statistical Association},volume={115},number={529},pages={455-466},year={2020},publisher={Taylor & Francis},doi={10.1080/01621459.2018.1562934},}
Conferences
2019
Inférence Variationnelle du Modèle à Blocs Stochastiques (SBM) avec covariables en présence de données manquantes
T. Tabouy, P. Barbillon, and J. Chiquet
In actes des 51\ieme journées françaises de statistique, Saclay, 2019
2018
Identifiabilité du modèle à Blocs Stochastiques en présence de données manquantes
T. Tabouy, P. Barbillon, and J. Chiquet
In actes des 50\ieme journées françaises de statistique, Saclay, 2018
2017
Inférence du Modèle à Blocs Stochastiques en présence de données manquantes
T. Tabouy, P. Barbillon, and J. Chiquet
In actes des 49\ieme journées françaises de statistique, Avignon, 2017