OpenMarkov, an Open-Source Tool for Probabilistic Graphical Models

Arias, Manuel; Pérez-Martín, Jorge; Luque, Manuel; Díez, F. J.

doi:10.24963/ijcai.2019/931

Cited by 11 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Further, these algorithms usually perform as black boxes, and it is difficult to determine to what degree those links are really supported by the data. Recently, Bermejo et al proposed an interactive learning technique which is implemented in OpenMarkov software. − This technique suggests a list of changes to improve the accuracy of the network but also allows an expert to intervene and select only those that make sense according to their domain knowledge. To extract the BN, we used the search and score method based on the K2 algorithm, , which is a greedy search technique with the following scoring metric where Đ is the data set of N samples each consisting of n variables; P ß( S ) ( Đ ) is the marginal likelihood of a specific network structure S given a data set of samples; N ij is the number of samples in the data set that have the j th value combination for the parents of the i th variable; and likewise N ijk is the number of samples with the i th variable in its k th states and its parents in their j th configuration.…”

Section: Machine Learning Methodsmentioning

confidence: 99%

Understanding and Predicting the Cause of Defects in Graphene Oxide Nanostructures Using Machine Learning

Motevalli

Sun

Barnard³

2020

J. Phys. Chem. C

View full text Add to dashboard Cite

Machine learning is a powerful way of uncovering hidden structure/property relationships in nanoscale materials, and it is tempting to assign structural causes to properties based on feature rankings reported by interpretable models. In this study of defective graphene oxide nanoflakes, we use classification, regression, and causal inference to show that not all important structural features directly influence the concentration of broken bonds, as a representative property. We find that while the presence of oxygen is important for actual bond breakage the presence and distribution of hydrogen determines how often bond breakage occurs.

show abstract

Section: Machine Learning Methodsmentioning

confidence: 99%

Understanding and Predicting the Cause of Defects in Graphene Oxide Nanostructures Using Machine Learning

Motevalli

Sun

Barnard³

2020

J. Phys. Chem. C

View full text Add to dashboard Cite

show abstract

“…Se realizó un análisis estadístico mediante los programas IBM SPSS Statistics 22 y OpenMarkov (Arias et al, 2019), así como una herramienta de código abierto y desarrollada por el Centro de Investigación sobre Sistemas Inteligentes de la UNED. En primer lugar, se efectuó un estudio descriptivo de las puntuaciones medias y un estudio correlacional en el que se analizó la fuerza de la relación entre las cinco primeras preguntas del cuestionario y la competencia digital y los tipos de formación realizada.…”

Section: Procedimiento De Recogida Y Análisis De Datosunclassified

Competencia digital de las personas orientadoras en los centros integrados de formación profesional de España

Requejo Fernández,

Raposo-Rivas,

Sarmiento Campos

2024

educar

View full text Add to dashboard Cite

La orientación profesional es clave para garantizar una educación de calidad e inclusiva y las tecnologías constituyen uno de los medios que pueden facilitar el alcance de ese desafío, por lo que son precisas ciertas habilidades en su manejo. Esta investigación tiene como objetivo conocer la competencia digital de las personas profesionales de la orientación en los centros de formación profesional de España. Se describen, se comparan y se analizan las respuestas emitidas por el 47,11% de la población a un cuestionario validado creado ad hoc que presenta una fiabilidad α = 0,81. Se desarrolla un estudio de datos descriptivo, correlacional y de redes bayesianas mediante los programas SPSS Statistics 22 y OpenMarkov. Los resultados afirman que la muestra posee una competencia digital a nivel usuario adquirida en su mayoría por autoformación y asistencia a conferencias. Además, los recursos y las herramientas más innovadoras están poco integrados en la orientación profesional, siendo las plataformas y los programas específicos de la orientación los más desconocidos. Como conclusión, las administraciones deben afrontar este desafío ofreciendo programas de formación permanente para mejorar las competencias digitales y, por ende, la integración de recursos tecnológicos como aliados para una educación más accesible e inclusiva.

show abstract

“…The table EU can be calculated with OpenMarkov (Arias and Díez, 2008;Arias et al, 2019;Díez et al, 2022), an open-source tool for probabilistic graphical models available at www.openmarkov.org. For these computations OpenMarkov only needs as input the ID model, including the graph (nodes and edges) and the numerical parameters.…”

Section: Computation Of the Expected Utility Tablementioning

confidence: 99%

Team assembly approach based in social modelling

Anaya

Letón

Luque

2023

Preprint

View full text Add to dashboard Cite

Team assembly is an important issue when collaborative or cooperative activities between people are carried out. In the educational context, it seems relevant to us to consider in the assembly process how students interact with others. Therefore, it is appropriate to carry out a process to model social behavior. In turn, the experts will be able to use the modeling information to establish the appropriate criteria to differentiate between assembly strategies according to the educational contexts. Our social behavior modeling process proposal is based on the information that can be extracted from the interactions and is guided by experts. We propose an approach divides into steps. Firstly, social network analysis metrics are used to study students’ social interactions. Secondly, a set of social interactions attributes are offered so that experts can organize them hierarchically. Thirdly a quantitative information on the students’ relationship that an influence diagram can infer due to the attributes hierarchically ordered. Once the modelling process is out, the qualitative information may be used to improve the team assembly. We describe in this paper two experiences where we have put it into practice our approach. The students made a peer review task. However, the pedagog-ical strategies that experts followed were different for each experience.

show abstract

OpenMarkov, an Open-Source Tool for Probabilistic Graphical Models

Cited by 11 publications

References 6 publications

Understanding and Predicting the Cause of Defects in Graphene Oxide Nanostructures Using Machine Learning

Understanding and Predicting the Cause of Defects in Graphene Oxide Nanostructures Using Machine Learning

Competencia digital de las personas orientadoras en los centros integrados de formación profesional de España

Team assembly approach based in social modelling

Contact Info

Product

Resources

About