Measuring the likelihood of models for network evolution

TL;DR

This paper introduces a framework to evaluate how well different models explain network evolution by calculating likelihoods, enabling comparison of hypotheses and incorporating linear combinations of models, tested on simulated and real networks.

Contribution

It presents a novel likelihood-based framework for assessing network evolution models, including methods for model comparison and optimization of model components.

Findings

Framework successfully distinguishes between different hypothesized models.

Method effectively optimizes component weights in linear model combinations.

Validated on both simulated and real network data.

Abstract

Many researchers have hypothesised models which explain the evolution of the topology of a target network. The framework described in this paper gives the likelihood that the target network arose from the hypothesised model. This allows rival hypothesised models to be compared for their ability to explain the target network. A null model (of random evolution) is proposed as a baseline for comparison. The framework also considers models made from linear combinations of model components. A method is given for the automatic optimisation of component weights. The framework is tested on simulated networks with known parameters and also on real data.