Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables
We propose and test improvements to state-of-the-art techniques of Bayeasian statistical inference based on pseudolikelihood maximization with $\ell_1$ regularization and with decimation. In particular, we present a method to determine the best value of the regularizer parameter starting from a...
Main Author: | |
---|---|
Format: | Article |
Language: | English |
Published: |
SciPost
2018-07-01
|
Series: | SciPost Physics |
Online Access: | https://scipost.org/SciPostPhys.5.1.002 |
id |
doaj-640b84d5cda040f99da65b599aa000fe |
---|---|
record_format |
Article |
spelling |
doaj-640b84d5cda040f99da65b599aa000fe2020-11-25T00:45:27ZengSciPostSciPost Physics2542-46532018-07-015100210.21468/SciPostPhys.5.1.002Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variablesAlessia Marruzzo, Payal Tyagi, Fabrizio Antenucci, Andrea Pagnani, Luca LeuzziWe propose and test improvements to state-of-the-art techniques of Bayeasian statistical inference based on pseudolikelihood maximization with $\ell_1$ regularization and with decimation. In particular, we present a method to determine the best value of the regularizer parameter starting from a hypothesis testing technique. Concerning the decimation, we also analyze the worst case scenario in which there is no sharp peak in the tilded-pseudolikelihood function, firstly defined as a criterion to stop the decimation. Techniques are applied to noisy systems with non-linear dynamics, mapped onto multi-variable interacting Hamiltonian effective models for waves and phasors. Results are analyzed varying the number of available samples and the externally tunable temperature-like parameter mimicing real data noise. Eventually the behavior of inference procedures described are tested against a wrong hypothesis: non-linearly generated data are analyzed with a pairwise interacting hypothesis. Our analysis shows that, looking at the behavior of the inverse graphical problem as data size increases, the methods exposed allow to rule out a wrong hypothesis.https://scipost.org/SciPostPhys.5.1.002 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Alessia Marruzzo, Payal Tyagi, Fabrizio Antenucci, Andrea Pagnani, Luca Leuzzi |
spellingShingle |
Alessia Marruzzo, Payal Tyagi, Fabrizio Antenucci, Andrea Pagnani, Luca Leuzzi Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables SciPost Physics |
author_facet |
Alessia Marruzzo, Payal Tyagi, Fabrizio Antenucci, Andrea Pagnani, Luca Leuzzi |
author_sort |
Alessia Marruzzo, Payal Tyagi, Fabrizio Antenucci, Andrea Pagnani, Luca Leuzzi |
title |
Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables |
title_short |
Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables |
title_full |
Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables |
title_fullStr |
Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables |
title_full_unstemmed |
Improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables |
title_sort |
improved pseudolikelihood regularization and decimation methods on non-linearly interacting systems with continuous variables |
publisher |
SciPost |
series |
SciPost Physics |
issn |
2542-4653 |
publishDate |
2018-07-01 |
description |
We propose and test improvements to state-of-the-art techniques of Bayeasian
statistical inference based on pseudolikelihood maximization with $\ell_1$
regularization and with decimation. In particular, we present a method to
determine the best value of the regularizer parameter starting from a
hypothesis testing technique. Concerning the decimation, we also analyze the
worst case scenario in which there is no sharp peak in the
tilded-pseudolikelihood function, firstly defined as a criterion to stop the
decimation. Techniques are applied to noisy systems with non-linear dynamics,
mapped onto multi-variable interacting Hamiltonian effective models for waves
and phasors. Results are analyzed varying the number of available samples and
the externally tunable temperature-like parameter mimicing real data noise.
Eventually the behavior of inference procedures described are tested against a
wrong hypothesis: non-linearly generated data are analyzed with a pairwise
interacting hypothesis. Our analysis shows that, looking at the behavior of the
inverse graphical problem as data size increases, the methods exposed allow to
rule out a wrong hypothesis. |
url |
https://scipost.org/SciPostPhys.5.1.002 |
work_keys_str_mv |
AT alessiamarruzzopayaltyagifabrizioantenucciandreapagnanilucaleuzzi improvedpseudolikelihoodregularizationanddecimationmethodsonnonlinearlyinteractingsystemswithcontinuousvariables |
_version_ |
1725270101394980864 |