Evaluación de los retrasos en actividades de construcción utilizando redes Bayesianas: Caso de estudio / Delay assessment of construction site activities using Bayesian networks: Case study / Avaliação dos atrasos nas atividades de construção utilizando redes Bayesianas: Estudo de caso

RESUMEN Los retrasos en proyectos de construcción son atribuídos a la concurrencia de factores que afectan el buen desarrollo, y mitigarlos, constituye uno de los mayores desafios que afronta la industria, ya que requiere tener en cuenta la dependencia e incidencia integrada de ellos. El objetivo de este estudio fue evaluar la influencia de factores de retraso sobre la duración de actividades de construcción, a través de un método que emplee las redes Bayesianas. Siguiendo la metodologia de investigación basada en el diseno, y aplicando un caso de estudio, se propone un método que incluye cómo identificar los principales factores de retraso que afecten las actividades; cómo representar la influencia y dependencia de estos factores por medio de una red Bayesiana; y cómo estimar el nivel de influencia por medio de la simulación de la red. Los resultados de esta investigación muestran como la aplicación de una red Bayesiana se puede emplear como apoyo a los profesionales de obra para gestionar las actividades de construcción y tomar decisiones en la medida que se actualice la información de la red bayesiana.

ABSTRACT The concurrence of multiple factors adversely affects construction project performance, resulting in delays. Mitigating such factors is challenging for the industry because solutions must include dependence and influence. Hence, managers must consider their systemic and integrated influence on the construction process for tracking projects. This study aimed to evaluate the influence of delay factors on the duration of construction site activities using Bayesian network techniques. Based on the design-based research methodology and the application in a case study this study proposed a method that involves three steps. First, identifying the main delay factors affecting construction activities; second, designing an influencing model as a Bayesian network; and third, estimating the integrated influence of such factors by simulating the Bayesian network. The results showed how a Bayesian network could support the construction team in managing the construction-site activities and making decisions about the performance of the construction process.

RESUMO Atrasos nos projetos de construção são atribuídos à concorrência de múltiplos fatores que afetam o desenvolvimento adequado do projeto, e mitigá-los é um dos maiores desafios enfrentados pelo setor; pois requer levar em conta a incidência integrada dos mesmos. O objetivo deste estudo foi avaliar a influência de um grupo de fatores sobre a duração das atividades de construção, utilizando a técnica da rede Bayesiana. Seguindo a metodologia de pesquisa baseada em projeto, foram identificados os principais fatores de atraso que afetam as atividades de construção e as relações de causa e efeito foram modeladas para estimar sua influência sobre a duração das atividades de fundação de um projeto de construção. Os resultados desta pesquisa mostram como a aplicação de uma rede Bayesiana pode ser usada para apoiar os profissionais do local no gerenciamento das atividades de construção e na tomada de decisões relativas ao desenvolvimento do projeto, considerando a incerteza e os fatores que influenciam o desenvolvimento do projeto.

Influence of Cushioning Variables in the Workplace and in the Family on the Probability of Suffering Stress

Stress at work and in the family is a very common issue in our society that generates many health-related problems. During recent years, numerous studies have sought to define the term stress, raising many contradictions that various authors have studied. Other authors have attempted to establish some criteria, in subjective and not very quantitative ways, in an attempt to reduce and even to eliminate stressors and their effects at work and in the family context. The purpose of this study was to quantify so-called cushioning variables, such as control, social support, home/work life conciliation, and even sports and leisure activities, with the purpose of, as much as possible, reducing the negative effects of stress, which seriously affects the health of workers. The study employs data from the Fifth European Working Conditions Survey, in which nearly 44,000 interviewees from 34 countries in the European Union participated. We constructed a probabilistic model based on a Bayesian network, using variables from both the workplace and the family, the aforementioned cushioning variables, as well as the variable stress. If action is taken on the above variables, then the probabilities of suffering high levels of stress may be reduced. Such action may improve the quality of life of people at work and in the family.

Precisión diagnóstica de un modelo de redes bayesianas en los síndromes coronarios agudos / Diagnostic accuracy of a bayesian network model in acute coronary syndromes

Introducción: La caracterización diagnóstica del dolor torácico, con énfasis en los síndromes coronarios agudos (SCA) es un requerimiento primordial para los médicos del área de urgencias. Objetivos: En el presente estudio se busca diseñar y evaluar el desempeño de las redes bayesianas en el apoyo al diagnóstico de los SCA. Metodología: Se trata de un estudio de pruebas diagnósticas en el cual se diseñaron dos modelos de redes bayesianas entrenadas en el framework OpenMarkov, a partir de las variables de la escala de probabilidad de Braunwald de angina en un grupo de 159 pacientes que luego se validó en una cohorte de 108 pacientes adultos hospitalizados con sospecha de un SCA en un hospital de tercer nivel de atención. Resultados: Se obtuvo una sensibilidad baja aunque con especificidad y valor predictivo positivo adecuados (62, 86 y 87% respectivamente). El rendimiento fue mejor en los casos que tuvieron electrocardiograma y biomarcadores negativos. Conclusiones: Un modelo de redes Bayesianas entrenado a partir de las variables de la escala de probabilidad de angina inestable de Braunwald, presenta un rendimiento aceptable para el diagnóstico de los SCA.

Introduction: The characterization and diagnosis of chest pain, with emphasis on acute coronary syndromes (ACS), is a fundamental requirement for the doctors at the emergency service. Objective: The aim of the present study is to design and evaluate the performance of Bayesian networks to back up the diagnosis of ACS. Methodology: A diagnostic tests study in which two models of Bayesian networks were designed and trained in the framework OpenMarkov, using the variables of the Braunwald angina probability scale in a group of 159 patients, which was validated afterwards in a cohort of 108 adult patients hospitalized with suspicion of ACS in a third level hospital. Results: Low sensitivity was obtained, with adequate specificity and positive predictive values, though (62, 86, and 87% respectively). Performance was better in the cases that had electrocardiogram and negative biomarkers. Conclusion: A model of Bayesian networks trained from the variables of the Braunwald unstable angina probability scale, exhibits an acceptable performance for the diagnosis of ACS.

Aprendizaje de clases de equivalencia de redes bayesianas basado en búsqueda competitiva de hormigas artificiales / Learning of bayesian networks equivalence classes based on competitive search of artificial ants

Este artículo propone un algoritmo de aprendizaje de clases de equivalencia de redes bayesianas basado en un algoritmo de búsqueda Greedy y modelos de búsqueda inspirados en hormigas competitivas. Específicamente para el algoritmo propuesto, se obtuvo una mejor aproximación entre la red predicha y la red bayesiana teórica de ejemplo ASIA, con respecto a algoritmos anteriores, para conjuntos de datos con 20 y 500 muestras. En promedio el algoritmo desarrollado obtuvo una aproximación con respecto a la distancia estructural de hamming de 10.7% y 5.3% menor comparada con la obtenida por los algoritmos Greedy y de colonia de hormigas (ACO-E) respectivamente para 20 muestras, y de hasta el 6.8% menor con respecto al algoritmo ACO-E para 500 muestras. Además, para 500 muestras el número de llamadas a la función de puntaje realizadas por el algoritmo propuesto fue menor que las realizadas por el algoritmo ACO-E en el 90% de las combinaciones, concluyendo que hubo una reducción de la complejidad computacional. Finalmente se presentan los resultados de la aplicación del algoritmo propuesto a un microarreglo obtenido por muestras de pacientes con Leucemia Mieloide Aguda (LMA) con 6 nuevas interacciones con dependencias estadísticas como potenciales interacciones biológicas con alta probabilidad.

This article proposes an algorithm for learning equivalence classes of Bayesian networks based on a Greed search algorithm and search patterns inspired by competitive ants. Specifically, for the proposed algorithm, we obtained a better approximation between the predicted network and the theoretical network ASIA with respect to previous algorithms for data sets with 20 and 500 samples. On average, the algorithm developed an approximation with respect to Structural Hamming Distance of 10.7% and 5.3% lower than Greedy algorithms and ACO-E respectively to 20 samples, and up to 6.8% lower tan ACO-E algorithm for 500 samples. Furthermore, for 500 samples the number of calls to the scoring function performed by the algorithm proposed was smaller than in the ACO-E algorithm in 90% of the combinations, concluding that there was a reduction in the computational complexity. Finally, we present the results of applying the proposed algorithm to a microarray samples obtained from patients with acute myeloid leukemia (AML) with 6 new interactions with statistical dependencies as potential biological interactions with high probability.

Using Bayesian Networks to Model Hierarchical Relationships in Epidemiological Studies / 한국역학회지

OBJECTIVES: To propose an alternative procedure, based on a Bayesian network (BN), for estimation and prediction, and to discuss its usefulness for taking into account the hierarchical relationships among covariates. METHODS: The procedure is illustrated by modeling the risk of diarrhea infection for 2,740 children aged 0 to 59 months in Cameroon. We compare the procedure with a standard logistic regression and with a model based on multi-level logistic regression. RESULTS: The standard logistic regression approach is inadequate, or at least incomplete, in that it does not attempt to account for potentially causal relationships between risk factors. The multi-level logistic regression does model the hierarchical structure, but does so in a piecewise manner; the resulting estimates and interpretations differ from those of the BN approach proposed here. An advantage of the BN approach is that it enables one to determine the probability that a risk factor (and/or the outcome) is in any specific state, given the states of the others. The currently available approaches can only predict the outcome (disease), given the states of the covariates. CONCLUSION: A major advantage of BNs is that they can deal with more complex interrelationships between variables whereas competing approaches deal at best only with hierarchical ones. We propose that BN be considered as well as a worthwhile method for summarizing the data in epidemiological studies whose aim is understanding the determinants of diseases and quantifying their effects.

Gene regulatory network models based on time series gene expression data: Recent progress / 第二军医大学学报

Regulation between genes is a dynamic event associated with changes of time and circumstances. Gene regulatory network is a complicated and dynamic system. Time series gene microarray provides a tool for creating dynamic gene regulatory network. In this paper,we review several models of dynamic gene regulatory network based on time series gene expression data, including temporal Boolean network,differential equation,dynamic Bayesian networks,etc.. The advantages and disadvantages of the models were analyzed and the future of the research is predicted.

Gene networks as a tool to understand transcriptional regulation

Gene regulatory networks, or simply gene networks (GNs), have shown to be a promising approach that the bioinformatics community has been developing for studying regulatory mechanisms in biological systems. GNs are built from the genome-wide high-throughput gene expression data that are often available from DNA microarray experiments. Conceptually, GNs are (un)directed graphs, where the nodes correspond to the genes and a link between a pair of genes denotes a regulatory interaction that occurs at transcriptional level. In the present study, we had two objectives: 1) to develop a framework for GN reconstruction based on a Bayesian network model that captures direct interactions between genes through nonparametric regression with B-splines, and 2) to demonstrate the potential of GNs in the analysis of expression data of a real biological system, the yeast pheromone response pathway. Our framework also included a number of search schemes to learn the network. We present an intuitive notion of GN theory as well as the detailed mathematical foundations of the model. A comprehensive analysis of the consistency of the model when tested with biological data was done through the analysis of the GNs inferred for the yeast pheromone pathway. Our results agree fairly well with what was expected based on the literature, and we developed some hypotheses about this system. Using this analysis, we intended to provide a guide on how GNs can be effectively used to study transcriptional regulation. We also discussed the limitations of GNs and the future direction of network analysis for genomic data. The software is available upon request.

Modeling of gene regulatory networks / 第二军医大学学报

Gene regulatory networks(GRN),which focuses on the complex interactions of genes in life,is an important part in the study of the functional genomics and is the frontier of bioinformatics research.Application of gene-chip technique in bioinformatics provides a great number of basic data for the research of GRN.This paper reviews the origin and recent development of GRN,explicates the preconditions and rationales for construction of GRN,and analyzes several classic GRN models: Boolean networks,linear models,non-linear models and Bayesian networks.The rationales,basic algorithms,advantages,disadvantages and applicability of the models are reviewed based on the characteristics of gene-chip data.