Inference on Complex Networks: From Structure to Dynamics
Associate Professor Marta Sales Pardo (Rovira i Virgili University, Spain)
Abstract: Bayesian inference is a robust mathematical framework to obtain plausible explanations of your data. A paramount aspect of Bayesian inference in the choice of probabilistic model to describe your data. I will talk about different modelling approaches to perform inference from complex network data. Specifically, I will discuss inference in two different problems: the inference of perturbations from partial observations, a problem which is of great importance in cellular biology in which you typically have access to limited information of the state of the cell; and the case of data aggregation from multiple layers - in particular I will discuss how to model multi-layer networks in a probabilistic way as a way to understand whether the typically aggregated networks we study have a multi-layer origin or not and what implications this has for network analysis run on those data.
Associate Professor Marta Sales Pardo (Rovira i Virgili University, Spain)
Abstract: Bayesian inference is a robust mathematical framework to obtain plausible explanations of your data. A paramount aspect of Bayesian inference in the choice of probabilistic model to describe your data. I will talk about different modelling approaches to perform inference from complex network data. Specifically, I will discuss inference in two different problems: the inference of perturbations from partial observations, a problem which is of great importance in cellular biology in which you typically have access to limited information of the state of the cell; and the case of data aggregation from multiple layers - in particular I will discuss how to model multi-layer networks in a probabilistic way as a way to understand whether the typically aggregated networks we study have a multi-layer origin or not and what implications this has for network analysis run on those data.
Home |
Calls |
Programme |
Registration |
Contact |
Copyright CCS © 2016