http://sedici.unlp.edu.ar:80/handle/10915/134656 Thu, 12 Nov 2026 07:45:41 GMT 2026-11-12T07:45:41Z http://sedici.unlp.edu.ar:80/handle/10915/135702 Contribucion a revista Electronic Journal of SADIO; vol. 16 Nota sobre el contenido del volumen 16 del Electronic Journal of SADIO. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135702 2017-06-04T00:00:00Z Nota sobre el contenido del volumen 16 del Electronic Journal of SADIO. http://sedici.unlp.edu.ar:80/handle/10915/135692 Articulo Electronic Journal of SADIO; vol. 16 La compatibilidad de un sitio web es la propiedad que lo permite ser apreciado idénticamente en todos los navegadores. La creciente necesidad de que los sitios web puedan ser compatibles, ha llevado a que las empresas realicen análisis de múltiples factores en materia de visibilidad durante las etapas de desarrollo. Pero la continua evolución de los navegadores hace que este análisis no sea suficiente para evitar incompatibilidades entre ellos, cobrando relevancia las pruebas de compatibilidad web. Por otro lado, hoy en día las empresas buscan alcanzar el enfoque de desarrollo de software continuo, donde el objetivo es mantener la producción de software en ciclos cortos de tiempo, acelerando los procesos de pruebas. Uno de los desafíos que existe actualmente, es la aceleración de las pruebas sobre la interfaz de usuario, entre ellas, las de compatibilidad. En este trabajo se presenta una técnica para automatizar la detección de defectos de incompatibilidad, mediante el uso del procesamiento digital de imágenes. Los resultados de su implementación indican que se adapta a los requerimientos de los procesos de desarrollo continuo, aumentando el rendimiento y la velocidad de las pruebas y disminuyendo la cantidad de falsos positivos que normalmente aparecen en este tipo de pruebas. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135692 2017-06-04T00:00:00Z La compatibilidad de un sitio web es la propiedad que lo permite ser apreciado idénticamente en todos los navegadores. La creciente necesidad de que los sitios web puedan ser compatibles, ha llevado a que las empresas realicen análisis de múltiples factores en materia de visibilidad durante las etapas de desarrollo. Pero la continua evolución de los navegadores hace que este análisis no sea suficiente para evitar incompatibilidades entre ellos, cobrando relevancia las pruebas de compatibilidad web. Por otro lado, hoy en día las empresas buscan alcanzar el enfoque de desarrollo de software continuo, donde el objetivo es mantener la producción de software en ciclos cortos de tiempo, acelerando los procesos de pruebas. Uno de los desafíos que existe actualmente, es la aceleración de las pruebas sobre la interfaz de usuario, entre ellas, las de compatibilidad. En este trabajo se presenta una técnica para automatizar la detección de defectos de incompatibilidad, mediante el uso del procesamiento digital de imágenes. Los resultados de su implementación indican que se adapta a los requerimientos de los procesos de desarrollo continuo, aumentando el rendimiento y la velocidad de las pruebas y disminuyendo la cantidad de falsos positivos que normalmente aparecen en este tipo de pruebas. http://sedici.unlp.edu.ar:80/handle/10915/135663 Articulo Electronic Journal of SADIO; vol. 16; https://sedici.unlp.edu.ar/handle/10915/57086 In this article we present an approach to dynamically validate the usage of software connectors in the context of software architectures. By employing aspect oriented techniques the system execution is monitored in order to obtain an architectural view describing how processes communicate and interact with each other. This output can later be compared to the connectors specied in the architecture document to validate the consistency between the architecture specication and the implementation of the system. Two case studies are presented showing the potential of the approach. We believe the results are promising enough to consider future extensions including other architectural elements beyond connectors. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135663 2017-06-04T00:00:00Z In this article we present an approach to dynamically validate the usage of software connectors in the context of software architectures. By employing aspect oriented techniques the system execution is monitored in order to obtain an architectural view describing how processes communicate and interact with each other. This output can later be compared to the connectors specied in the architecture document to validate the consistency between the architecture specication and the implementation of the system. Two case studies are presented showing the potential of the approach. We believe the results are promising enough to consider future extensions including other architectural elements beyond connectors. http://sedici.unlp.edu.ar:80/handle/10915/135161 Articulo VIII Congreso Argentino de AgroInformática (CAI-2016) - JAIIO 45 (Tres de Febrero, 2016); Electronic Journal of SADIO; vol. 16; https://sedici.unlp.edu.ar/handle/10915/57431 Evapotranspiration is an important component of hydrologic balance and represent essential information for irrigation scheduling and water resources planning. Sometimes, the use of the recommended Penman-Monteith method is restricted by the lack of input variables and, therefore, empirical methods become essential.The study aimed: a) to develop and evaluate the performance of models based on artificial neural networks (ANN) to estimate daily values of reference evapotranspiration (ET0PM) with a limited number of input variables and b)to apply methods of knowledge extraction based on connection weights and sensitivity analysis to better understanding of ANN. Daily evapotranspiration values computed following the Penman-Monteith equation (ET0PM), were used as target outputs for the implementation of the ANN. Data of global radiation (Rg), net radiation (Rn) and extraterrestrial radiation (RTA) were alternated in combinations with air temperature (Ta), vapor pressure deficit (DPV) and wind (u) as inputs to networksof type multilayer perceptron. Also, combinations with basis in RTA and minimum and maximum air temperatures(Tmin, Tmax) were tested. The ANN with best performance for each combination of inputs were retained to evaluate the performance based on multi-criteria analysis. According to the results, it can be concluded that it is possible to estimate accurately daily ET0PM values. Air temperature and deficit of pressure vapor were found to be more effective than wind velocity in modelling ET0, whichever the radiation (Rn, Rg or RTA) used as input.A decomposition method based on Garson’s algorithm was applied to quantify the relative importance for each input variable. Sensitivity analysis was also performed to identify relevant inputs and quantify the risk of a certain combination of inputs on target values. The application of complementary proceduresin evaluation of ANN models is discussed, paying attentionespecially on detection of the better predicting variables and analysis of errors. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135161 2017-06-04T00:00:00Z Evapotranspiration is an important component of hydrologic balance and represent essential information for irrigation scheduling and water resources planning. Sometimes, the use of the recommended Penman-Monteith method is restricted by the lack of input variables and, therefore, empirical methods become essential.The study aimed: a) to develop and evaluate the performance of models based on artificial neural networks (ANN) to estimate daily values of reference evapotranspiration (ET0PM) with a limited number of input variables and b)to apply methods of knowledge extraction based on connection weights and sensitivity analysis to better understanding of ANN. Daily evapotranspiration values computed following the Penman-Monteith equation (ET0PM), were used as target outputs for the implementation of the ANN. Data of global radiation (Rg), net radiation (Rn) and extraterrestrial radiation (RTA) were alternated in combinations with air temperature (Ta), vapor pressure deficit (DPV) and wind (u) as inputs to networksof type multilayer perceptron. Also, combinations with basis in RTA and minimum and maximum air temperatures(Tmin, Tmax) were tested. The ANN with best performance for each combination of inputs were retained to evaluate the performance based on multi-criteria analysis. According to the results, it can be concluded that it is possible to estimate accurately daily ET0PM values. Air temperature and deficit of pressure vapor were found to be more effective than wind velocity in modelling ET0, whichever the radiation (Rn, Rg or RTA) used as input.A decomposition method based on Garson’s algorithm was applied to quantify the relative importance for each input variable. Sensitivity analysis was also performed to identify relevant inputs and quantify the risk of a certain combination of inputs on target values. The application of complementary proceduresin evaluation of ANN models is discussed, paying attentionespecially on detection of the better predicting variables and analysis of errors. http://sedici.unlp.edu.ar:80/handle/10915/135160 Articulo Electronic Journal of SADIO; vol. 16 We introduce a model for the spatial distribution of cell phone antennas in the city of Buenos Aires (CABA). After showing that the complete spatial randomness (homogeneous Poisson distribution) hypothesis does not hold, we propose a model in which each point is distributed according to a bivariate Gaussian variable with mean given by the barycenter of its neighbors in the Delaunay triangulation. We show that this model is suitable, and can be used to generate a synthetic distribution of antennas. The generated distribution contains no sensitive or proprietary information, and can thus be freely shared with research groups, fostering further research on the subject. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135160 2017-06-04T00:00:00Z We introduce a model for the spatial distribution of cell phone antennas in the city of Buenos Aires (CABA). After showing that the complete spatial randomness (homogeneous Poisson distribution) hypothesis does not hold, we propose a model in which each point is distributed according to a bivariate Gaussian variable with mean given by the barycenter of its neighbors in the Delaunay triangulation. We show that this model is suitable, and can be used to generate a synthetic distribution of antennas. The generated distribution contains no sensitive or proprietary information, and can thus be freely shared with research groups, fostering further research on the subject. http://sedici.unlp.edu.ar:80/handle/10915/135159 Articulo Electronic Journal of SADIO; vol. 16 In this work an iterative method to solve the nonlinear least squares problem is presented. The algorithm combines a secant method with a strategy of nonmonotone trust region. In order to dene the quadratic model, the Hessian matrix is chosen using a secant approach that takes advantage of the structure of the problem, and the radius of the trust region is updated following an adaptive technique. Moreover, convergence properties of this algorithm are proved. The numerical experimentation, in which several ways of choosing the Hessian matrix are compared, shows the effiency and robustness of the method. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135159 2017-06-04T00:00:00Z In this work an iterative method to solve the nonlinear least squares problem is presented. The algorithm combines a secant method with a strategy of nonmonotone trust region. In order to dene the quadratic model, the Hessian matrix is chosen using a secant approach that takes advantage of the structure of the problem, and the radius of the trust region is updated following an adaptive technique. Moreover, convergence properties of this algorithm are proved. The numerical experimentation, in which several ways of choosing the Hessian matrix are compared, shows the effiency and robustness of the method. http://sedici.unlp.edu.ar:80/handle/10915/135156 Articulo Electronic Journal of SADIO; vol. 16 We (the machine learning team at Jampp) strive to predict click-through rates (CTR) and conversion rates (CVR) for the real-time bidding (RTB) online advertising market by means of an in-house online machine learning platform based on a state-of-the-art stochastic gradient descent estimator. Our estimation framework has already been covered in a previous paper, so here we want to focus on some peripheral aspects of our platform that, in spite of being of a somewhat ancillary nature, nevertheless tend to dominate development efforts and overall system complexity; namely, in order to feed the learning system we first need to sample a very high-volume stream of out-of-order and scattered-in-time events and consolidate them into a sequence of observations representing the underlying market transactions, each observation composed of a set of features and a response, from which the estimator is ultimately able to learn. This paper is written in a down-to-earth fashion: we describe a number of particular difficulties the general problem of sampling in an online high-volume setting poses and then we present our concrete answers to those difficulties based on real, hands-on, experience. Sun, 04 Jun 2017 00:00:00 GMT http://sedici.unlp.edu.ar:80/handle/10915/135156 2017-06-04T00:00:00Z We (the machine learning team at Jampp) strive to predict click-through rates (CTR) and conversion rates (CVR) for the real-time bidding (RTB) online advertising market by means of an in-house online machine learning platform based on a state-of-the-art stochastic gradient descent estimator. Our estimation framework has already been covered in a previous paper, so here we want to focus on some peripheral aspects of our platform that, in spite of being of a somewhat ancillary nature, nevertheless tend to dominate development efforts and overall system complexity; namely, in order to feed the learning system we first need to sample a very high-volume stream of out-of-order and scattered-in-time events and consolidate them into a sequence of observations representing the underlying market transactions, each observation composed of a set of features and a response, from which the estimator is ultimately able to learn. This paper is written in a down-to-earth fashion: we describe a number of particular difficulties the general problem of sampling in an online high-volume setting poses and then we present our concrete answers to those difficulties based on real, hands-on, experience.