2025
Pérez-Vieites, Sara; Molina-Bulla, Harold; Míguez, Joaquín
Nested smoothing algorithms for inference and tracking of heterogeneous multi-scale state-space systems Artículo de revista
En: Foundations of Data Science, pp. 0–0, 2025, (Publisher: Foundations of Data Science).
@article{perez-vieites_nested_2025,
title = {Nested smoothing algorithms for inference and tracking of heterogeneous multi-scale state-space systems},
author = {Sara Pérez-Vieites and Harold Molina-Bulla and Joaquín Míguez},
url = {https://www.aimsciences.org/en/article/doi/10.3934/fods.2025002},
doi = {10.3934/fods.2025002},
year = {2025},
date = {2025-02-01},
urldate = {2025-02-27},
journal = {Foundations of Data Science},
pages = {0–0},
abstract = {Multi-scale problems, where variables of interest evolve in different time-scales and live in different state-spaces, can be found in many fields of science. Here, we introduce a new recursive methodology for Bayesian inference that aims at estimating the static parameters and tracking the dynamic variables of these kind of systems. Although the proposed approach works in rather general setups, for clarity we analyze the case of a heterogeneous multi-scale model with 3 time-scales (static parameters, slow dynamic state variables and fast dynamic state variables). The proposed scheme, based on a nested filtering methodology of [27], combines three intertwined layers of filtering techniques that approximate recursively the joint posterior probability distribution of the parameters and both sets of dynamic state variables given a sequence of partial and noisy observations. We explore the use of both sequential Monte Carlo schemes and several Kalman filtering techniques in the different layers of the methodology to obtain approximations of the posterior probability distributions of interest. Some numerical results are presented for a stochastic two-scale Lorenz 96 model with unknown parameters.},
note = {Publisher: Foundations of Data Science},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Multi-scale problems, where variables of interest evolve in different time-scales and live in different state-spaces, can be found in many fields of science. Here, we introduce a new recursive methodology for Bayesian inference that aims at estimating the static parameters and tracking the dynamic variables of these kind of systems. Although the proposed approach works in rather general setups, for clarity we analyze the case of a heterogeneous multi-scale model with 3 time-scales (static parameters, slow dynamic state variables and fast dynamic state variables). The proposed scheme, based on a nested filtering methodology of [27], combines three intertwined layers of filtering techniques that approximate recursively the joint posterior probability distribution of the parameters and both sets of dynamic state variables given a sequence of partial and noisy observations. We explore the use of both sequential Monte Carlo schemes and several Kalman filtering techniques in the different layers of the methodology to obtain approximations of the posterior probability distributions of interest. Some numerical results are presented for a stochastic two-scale Lorenz 96 model with unknown parameters.
Sztamfater-Garcia, Yannick; Sanjurjo-Rivo, Manuel; Escribano, Guillermo; Molina-Bulla, Harold; Miguez, Joaquin
An approximate model for the computation of in-orbit collision probabilities using importance sampling Artículo de revista
En: Advances in Space Research, 2025, ISSN: 0273-1177.
@article{sztamfater-garcia_approximate_2025,
title = {An approximate model for the computation of in-orbit collision probabilities using importance sampling},
author = {Yannick Sztamfater-Garcia and Manuel Sanjurjo-Rivo and Guillermo Escribano and Harold Molina-Bulla and Joaquin Miguez},
url = {https://www.sciencedirect.com/science/article/pii/S0273117724013206},
doi = {10.1016/j.asr.2024.12.074},
issn = {0273-1177},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-16},
journal = {Advances in Space Research},
abstract = {The risk of orbital collisions is at an all-time high. Standard methods which compute the probability of collision (PoC) often make restrictive assumptions which produce numerically efficient solutions, the quality of which may decrease over long time-spans. The crude Monte Carlo (CMC) simulation makes few assumptions and can produce high quality estimates; however, its computational cost can be prohibitively high. We introduce a new method based on importance sampling (IS), which aims to attain the accuracy of a CMC simulation at a fraction of the computational cost. To accomplish this, we propose a surrogate dynamical model which exploits Keplerian conjunction geometry and applies a linear correction to approximate the full-dynamics of the conjunction scenario, reducing the computational load of evaluating the dynamical model. This linear approximation is satisfactorily validated by means of high-fidelity model evaluations. The PoC results, when the method is applied on LEO and GEO scenarios, show agreement with an independent CMC benchmark simulation for the calculation of the PoC at a fraction of the computational cost.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The risk of orbital collisions is at an all-time high. Standard methods which compute the probability of collision (PoC) often make restrictive assumptions which produce numerically efficient solutions, the quality of which may decrease over long time-spans. The crude Monte Carlo (CMC) simulation makes few assumptions and can produce high quality estimates; however, its computational cost can be prohibitively high. We introduce a new method based on importance sampling (IS), which aims to attain the accuracy of a CMC simulation at a fraction of the computational cost. To accomplish this, we propose a surrogate dynamical model which exploits Keplerian conjunction geometry and applies a linear correction to approximate the full-dynamics of the conjunction scenario, reducing the computational load of evaluating the dynamical model. This linear approximation is satisfactorily validated by means of high-fidelity model evaluations. The PoC results, when the method is applied on LEO and GEO scenarios, show agreement with an independent CMC benchmark simulation for the calculation of the PoC at a fraction of the computational cost.
2024
Miguez, Joaquín; Molina-Bulla, Harold; Mariño, Inés P.
Master-slave coupling scheme for synchronization and parameter estimation in the generalized Kuramoto-Sivashinsky equation Artículo de revista
En: Physical Review E, vol. 110, no 5, pp. 054206, 2024, (Publisher: American Physical Society).
@article{miguez_master-slave_2024-1,
title = {Master-slave coupling scheme for synchronization and parameter estimation in the generalized Kuramoto-Sivashinsky equation},
author = {Joaquín Miguez and Harold Molina-Bulla and Inés P. Mariño},
url = {https://link.aps.org/doi/10.1103/PhysRevE.110.054206},
doi = {10.1103/PhysRevE.110.054206},
year = {2024},
date = {2024-11-01},
urldate = {2024-11-06},
journal = {Physical Review E},
volume = {110},
number = {5},
pages = {054206},
abstract = {The problem of estimating the constant parameters of the Kuramoto-Sivashinsky (KS) equation from observed data has received attention from researchers in physics, applied mathematics, and statistics. This is motivated by the various physical applications of the equation and also because it often serves as a test model for the study of space-time pattern formation. Remarkably, most existing inference techniques rely on statistical tools, which are computationally very costly yet do not exploit the dynamical features of the system. In this paper, we introduce a simple, online parameter estimation method that relies on the synchronization properties of the KS equation. In particular, we describe a master-slave setup where the slave model is driven by observations from the master system. The slave dynamics are data-driven and designed to continuously adapt the model parameters until identical synchronization with the master system is achieved. We provide a simple analysis that supports the proposed approach and also present and discuss the results of an extensive set of computer simulations. Our numerical study shows that the proposed method is computationally fast and also robust to initialization errors, observational noise, and variations in the spatial resolution of the numerical scheme used to integrate the KS equation.},
note = {Publisher: American Physical Society},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The problem of estimating the constant parameters of the Kuramoto-Sivashinsky (KS) equation from observed data has received attention from researchers in physics, applied mathematics, and statistics. This is motivated by the various physical applications of the equation and also because it often serves as a test model for the study of space-time pattern formation. Remarkably, most existing inference techniques rely on statistical tools, which are computationally very costly yet do not exploit the dynamical features of the system. In this paper, we introduce a simple, online parameter estimation method that relies on the synchronization properties of the KS equation. In particular, we describe a master-slave setup where the slave model is driven by observations from the master system. The slave dynamics are data-driven and designed to continuously adapt the model parameters until identical synchronization with the master system is achieved. We provide a simple analysis that supports the proposed approach and also present and discuss the results of an extensive set of computer simulations. Our numerical study shows that the proposed method is computationally fast and also robust to initialization errors, observational noise, and variations in the spatial resolution of the numerical scheme used to integrate the KS equation.
2018
Bousono-Calzon, Carlos; Molina-Bulla, Harold; Escudero-Garzas, Jose Joaquin; Herrera-Galvez, Francisco J.
Expert Selection in Prediction Markets With Homological Invariants Artículo de revista
En: IEEE Access, vol. 6, pp. 32226–32239, 2018, ISSN: 2169-3536.
@article{bousono-calzon_expert_2018,
title = {Expert Selection in Prediction Markets With Homological Invariants},
author = {Carlos Bousono-Calzon and Harold Molina-Bulla and Jose Joaquin Escudero-Garzas and Francisco J. Herrera-Galvez},
url = {https://ieeexplore.ieee.org/document/8386656/},
doi = {10.1109/ACCESS.2018.2846878},
issn = {2169-3536},
year = {2018},
date = {2018-01-01},
urldate = {2023-07-31},
journal = {IEEE Access},
volume = {6},
pages = {32226–32239},
abstract = {Group decision making is a topic of growing interest in today’s complex societies. One of the key technologies in this area is the prediction market, where a group of experts plays a fake stock market with assets that represent the outcomes of an uncertain event. The particular problem we address in this paper is the expert selection in these markets to improve their reliability. To aggregate decisions from a particular group of experts, instead of using prices as is typically done, we define a market deconstruction considering player portfolios. This decision technology makes the behaviors of experts toward their decisions available through their portfolios evolution. Our main contribution is the identification of two Persistent Homological Invariants able to classify experts in groups based on the histories of their portfolios. Interestingly, this translates into the definition of essentially two dominant groups. A simulation of the Prediction Market with artificial agents allow us to interpret these two classes as rational and irrational players, following the Microeconomic jargon. Four experiments with experts in the insurance sector help us to illustrate the relationship between these two player types with the prediction reliability of the market.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Group decision making is a topic of growing interest in today’s complex societies. One of the key technologies in this area is the prediction market, where a group of experts plays a fake stock market with assets that represent the outcomes of an uncertain event. The particular problem we address in this paper is the expert selection in these markets to improve their reliability. To aggregate decisions from a particular group of experts, instead of using prices as is typically done, we define a market deconstruction considering player portfolios. This decision technology makes the behaviors of experts toward their decisions available through their portfolios evolution. Our main contribution is the identification of two Persistent Homological Invariants able to classify experts in groups based on the histories of their portfolios. Interestingly, this translates into the definition of essentially two dominant groups. A simulation of the Prediction Market with artificial agents allow us to interpret these two classes as rational and irrational players, following the Microeconomic jargon. Four experiments with experts in the insurance sector help us to illustrate the relationship between these two player types with the prediction reliability of the market.
2014
Barreto-Rodriguez, Carol Milena; Ramirez-Angulo, Jessica Paola; Ramirez, Jorge Mario Gomez; Achenie, LUKE; Molina-Bulla, Harold; Barrios, Andrés Fernando González
Dynamic Flux Balance Analysis for Predicting Gene Overexpression Effects in Batch Cultures Artículo de revista
En: Journal of Biological Systems, vol. 22, pp. 327–328, 2014, ISSN: 0218-3390, (Publisher: World Scientific Publishing Company).
@article{barreto-rodriguez_dynamic_2014,
title = {Dynamic Flux Balance Analysis for Predicting Gene Overexpression Effects in Batch Cultures},
author = {Carol Milena Barreto-Rodriguez and Jessica Paola Ramirez-Angulo and Jorge Mario Gomez Ramirez and LUKE Achenie and Harold Molina-Bulla and Andrés Fernando González Barrios},
url = {https://www.worldscientific.com/doi/epdf/10.1142/S0218339014500107},
doi = {10.1142/S0218339014500107},
issn = {0218-3390},
year = {2014},
date = {2014-08-01},
urldate = {2023-07-31},
journal = {Journal of Biological Systems},
volume = {22},
pages = {327–328},
abstract = {The advent of numerous technological platforms for genome sequencing has led to increasing understanding and construction of metabolic networks. A popular system engineering strategy is used to ana...},
note = {Publisher: World Scientific Publishing Company},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The advent of numerous technological platforms for genome sequencing has led to increasing understanding and construction of metabolic networks. A popular system engineering strategy is used to ana...
2011
Gomez-Verdejo, Vanessa; Martinez-Ramon, Manel; Arenas-Garcia, Jerónimo; Lazaro-Gredilla, Miguel; Molina-Bulla, Harold
Support Vector Machines With Constraints for Sparsity in the Primal Parameters Artículo de revista
En: IEEE Transactions on Neural Networks, vol. 22, no 8, pp. 1269–1283, 2011, ISSN: 1941-0093, (Conference Name: IEEE Transactions on Neural Networks).
@article{gomez-verdejo_support_2011,
title = {Support Vector Machines With Constraints for Sparsity in the Primal Parameters},
author = {Vanessa Gomez-Verdejo and Manel Martinez-Ramon and Jerónimo Arenas-Garcia and Miguel Lazaro-Gredilla and Harold Molina-Bulla},
doi = {10.1109/TNN.2011.2148727},
issn = {1941-0093},
year = {2011},
date = {2011-08-01},
journal = {IEEE Transactions on Neural Networks},
volume = {22},
number = {8},
pages = {1269–1283},
abstract = {This paper introduces a new support vector machine (SVM) formulation to obtain sparse solutions in the primal SVM parameters, providing a new method for feature selection based on SVMs. This new approach includes additional constraints to the classical ones that drop the weights associated to those features that are likely to be irrelevant. A ν-SVM formulation has been used, where ν indicates the fraction of features to be considered. This paper presents two versions of the proposed sparse classifier, a 2-norm SVM and a 1-norm SVM, the latter having a reduced computational burden with respect to the first one. Additionally, an explanation is provided about how the presented approach can be readily extended to multiclass classification or to problems where groups of features, rather than isolated features, need to be selected. The algorithms have been tested in a variety of synthetic and real data sets and they have been compared against other state of the art SVM-based linear feature selection methods, such as 1-norm SVM and doubly regularized SVM. The results show the good feature selection ability of the approaches.},
note = {Conference Name: IEEE Transactions on Neural Networks},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper introduces a new support vector machine (SVM) formulation to obtain sparse solutions in the primal SVM parameters, providing a new method for feature selection based on SVMs. This new approach includes additional constraints to the classical ones that drop the weights associated to those features that are likely to be irrelevant. A ν-SVM formulation has been used, where ν indicates the fraction of features to be considered. This paper presents two versions of the proposed sparse classifier, a 2-norm SVM and a 1-norm SVM, the latter having a reduced computational burden with respect to the first one. Additionally, an explanation is provided about how the presented approach can be readily extended to multiclass classification or to problems where groups of features, rather than isolated features, need to be selected. The algorithms have been tested in a variety of synthetic and real data sets and they have been compared against other state of the art SVM-based linear feature selection methods, such as 1-norm SVM and doubly regularized SVM. The results show the good feature selection ability of the approaches.
2004
Pedraza-Jimenez, Rafael; Valverde-Albacete, Francisco J.; Molina-Bulla, Harold; Cid-Sueiro, Jesús; Navia-Vazquez, Angel
Assessment and Reuse of Contents in the Competence-Based Educational Platform InterMediActor Artículo de revista
En: WSEAS Trans. On Computers, vol. 3, num.1, pp. 115–121, 2004.
@article{pedraza-jimenez_assessment_2004,
title = {Assessment and Reuse of Contents in the Competence-Based Educational Platform InterMediActor},
author = {Rafael Pedraza-Jimenez and Francisco J. Valverde-Albacete and Harold Molina-Bulla and Jesús Cid-Sueiro and Angel Navia-Vazquez},
year = {2004},
date = {2004-01-01},
journal = {WSEAS Trans. On Computers},
volume = {3, num.1},
pages = {115–121},
abstract = {This paper describes a failure alert system and a methodology for content reuse in a new instructional design system called InterMediActor (IMA). IMA provides an environment for instructional content design, production and reuse, and for students’ evaluation based in content specification through a hierarchical structure of competences. The student assessment process and information extraction process for content reuse are explained.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper describes a failure alert system and a methodology for content reuse in a new instructional design system called InterMediActor (IMA). IMA provides an environment for instructional content design, production and reuse, and for students’ evaluation based in content specification through a hierarchical structure of competences. The student assessment process and information extraction process for content reuse are explained.
2003
Valverde-Albacete, Francisco J.; Pedraza-Jimenez, Rafael; Molina-Bulla, Harold; Cid-Sueiro, Jesús; Díaz, Paloma; Navia-Vazquez, Angel
InterMediActor: an Environment for Instructional Content Design Based on Competences. Artículo de revista
En: Journal of Educational Technology & Society, vol. 6, no 4, pp. 575–579, 2003, ISSN: 1436-4522.
@article{valverde-albacete_intermediactor_2003,
title = {InterMediActor: an Environment for Instructional Content Design Based on Competences.},
author = {Francisco J. Valverde-Albacete and Rafael Pedraza-Jimenez and Harold Molina-Bulla and Jesús Cid-Sueiro and Paloma Díaz and Angel Navia-Vazquez},
issn = {1436-4522},
year = {2003},
date = {2003-10-01},
journal = {Journal of Educational Technology & Society},
volume = {6},
number = {4},
pages = {575–579},
keywords = {},
pubstate = {published},
tppubtype = {article}
}