Network reconstruction via the minimum description length principle

Tiago P. Peixoto

2024-05-08

Abstract:A fundamental problem associated with the task of network reconstruction from dynamical or behavioral data consists in determining the most appropriate model complexity in a manner that prevents overfitting, and produces an inferred network with a statistically justifiable number of edges. The status quo in this context is based on $L_{1}$ regularization combined with cross-validation. However, besides its high computational cost, this commonplace approach unnecessarily ties the promotion of sparsity with weight "shrinkage". This combination forces a trade-off between the bias introduced by shrinkage and the network sparsity, which often results in substantial overfitting even after cross-validation. In this work, we propose an alternative nonparametric regularization scheme based on hierarchical Bayesian inference and weight quantization, which does not rely on weight shrinkage to promote sparsity. Our approach follows the minimum description length (MDL) principle, and uncovers the weight distribution that allows for the most compression of the data, thus avoiding overfitting without requiring cross-validation. The latter property renders our approach substantially faster to employ, as it requires a single fit to the complete data. As a result, we have a principled and efficient inference scheme that can be used with a large variety of generative models, without requiring the number of edges to be known in advance. We also demonstrate that our scheme yields systematically increased accuracy in the reconstruction of both artificial and empirical networks. We highlight the use of our method with the reconstruction of interaction networks between microbial communities from large-scale abundance samples involving in the order of $10^{4}$ to $10^{5}$ species, and demonstrate how the inferred model can be used to predict the outcome of interventions in the system.

Machine Learning,Social and Information Networks,Data Analysis, Statistics and Probability,Populations and Evolution

What problem does this paper attempt to address?

The paper aims to address the issue of overfitting in the network reconstruction process. The current commonly used method is L1 regularization, which has the disadvantages of requiring preset sparsity, weight shrinkage leading to bias, and the need to determine parameters through cross-validation. The paper proposes a new regularization scheme based on hierarchical Bayesian inference and weight quantization, following the principle of minimum description length. This scheme can compress data without relying on cross-validation, improve reconstruction accuracy, and is suitable for various generative models, with no strong assumption on weight distribution. This method is more effective and concise, especially suitable for the reconstruction of large networks.