Journal
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT
Volume 2022, Issue 5, Pages -Publisher
IOP Publishing Ltd
DOI: 10.1088/1742-5468/ac650c
Keywords
inference of graphical models; statistical inference in biological systems; protein function and design
Categories
Ask authors/readers for more resources
Understanding the role of regularization in statistical inference is important, and well-chosen regularization schemes can significantly improve the quality of inferred models. This study focuses on the L-2 regularization in maximum a posteriori (MAP) inference of generative pairwise graphical models. Analytical calculations and numerical experiments are used to examine the likelihoods of different data sets under various regularization strengths. The results show that the test and generated likelihoods have close values at their maximum, and the optimal regularization strength is related to the magnitude of couplings in the underlying network.
Understanding the role of regularization is a central question in statistical inference. Empirically, well-chosen regularization schemes often dramatically improve the quality of the inferred models by avoiding overfitting of the training data. We consider here the particular case of L-2 regularization in the maximum a posteriori (MAP) inference of generative pairwise graphical models. Based on analytical calculations on Gaussian multivariate distributions and numerical experiments on Gaussian and Potts models we study the likelihoods of the training, test, and 'generated data' (with the inferred models) sets as functions of the regularization strengths. We show in particular that, at its maximum, the test likelihood and the 'generated' likelihood, which quantifies the quality of the generated samples, have remarkably close values. The optimal value for the regularization strength is found to be approximately equal to the inverse sum of the squared couplings incoming on sites on the underlying network of interactions. Our results seem to be robust against changes in the structure of the ground-truth underlying interactions that generated the data, when small fluctuations of the posterior distribution around the MAP estimator are taken into account, and when L-1 regularization is considered (instead of L-2). Connections with empirical works on protein models learned from homologous sequences are discussed.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available