Examinando por Materia "Learning systems"
Mostrando 1 - 14 de 14
Resultados por página
Opciones de ordenación
Ítem Automatic detection of building typology using deep learning methods on street level images(PERGAMON-ELSEVIER SCIENCE LTD, 2020-03-20) Duque, J.; Gonzalez, D.; Rueda Plata, Diego; Acevedo, A.; Ramos, R.; Betancourt, A.; García, S.; Mecánica AplicadaAn exposure model is a key component for assessing potential human and economic losses from natural disasters. An exposure model consists of a spatially disaggregated description of the infrastructure and population of a region under study. Depending on the size of the settlement area, developing such models can be a costly and time-consuming task. In this paper we use a manually annotated dataset consisting of approximately 10,000 photos acquired at street level in the urban area of Medellín to explore the potential for using a convolutional neural network (CNN) to automatically detect building materials and types of lateral-load resisting systems, which are attributes that define a building's structural typology (which is a key issue in exposure models for seismic risk assessment). The results of the developed model achieved a precision of 93% and a recall of 95% when identifying nonductile buildings, which are the buildings most likely to be damaged in an earthquake. Identifying fine-grained material typology is more difficult, because many visual clues are physically hidden, but our model matches expert level performances, achieving a recall of 85% and accuracy scores ranging from 60% to 82% on the three most common building typologies, which account for 91% of the total building population in Medellín. Overall, this study shows that a CNN can make a substantial contribution to developing cost-effective exposure models. © 2020 Elsevier LtdÍtem Automatic detection of building typology using deep learning methods on street level images(PERGAMON-ELSEVIER SCIENCE LTD, 2020-03-20) Duque, J.; Gonzalez, D.; Rueda Plata, Diego; Acevedo, A.; Ramos, R.; Betancourt, A.; García, S.; Universidad EAFIT. Departamento de Economía y Finanzas; Research in Spatial Economics (RISE)An exposure model is a key component for assessing potential human and economic losses from natural disasters. An exposure model consists of a spatially disaggregated description of the infrastructure and population of a region under study. Depending on the size of the settlement area, developing such models can be a costly and time-consuming task. In this paper we use a manually annotated dataset consisting of approximately 10,000 photos acquired at street level in the urban area of Medellín to explore the potential for using a convolutional neural network (CNN) to automatically detect building materials and types of lateral-load resisting systems, which are attributes that define a building's structural typology (which is a key issue in exposure models for seismic risk assessment). The results of the developed model achieved a precision of 93% and a recall of 95% when identifying nonductile buildings, which are the buildings most likely to be damaged in an earthquake. Identifying fine-grained material typology is more difficult, because many visual clues are physically hidden, but our model matches expert level performances, achieving a recall of 85% and accuracy scores ranging from 60% to 82% on the three most common building typologies, which account for 91% of the total building population in Medellín. Overall, this study shows that a CNN can make a substantial contribution to developing cost-effective exposure models. © 2020 Elsevier LtdÍtem Automatic detection of building typology using deep learning methods on street level images(PERGAMON-ELSEVIER SCIENCE LTD, 2020-03-20) Duque, J.; Gonzalez, D.; Rueda Plata, Diego; Acevedo, A.; Ramos, R.; Betancourt, A.; García, S.; Duque, J.; Gonzalez, D.; Rueda Plata, Diego; Acevedo, A.; Ramos, R.; Betancourt, A.; García, S.; Universidad EAFIT. Departamento de Ingeniería de Producción; Materiales de IngenieríaAn exposure model is a key component for assessing potential human and economic losses from natural disasters. An exposure model consists of a spatially disaggregated description of the infrastructure and population of a region under study. Depending on the size of the settlement area, developing such models can be a costly and time-consuming task. In this paper we use a manually annotated dataset consisting of approximately 10,000 photos acquired at street level in the urban area of Medellín to explore the potential for using a convolutional neural network (CNN) to automatically detect building materials and types of lateral-load resisting systems, which are attributes that define a building's structural typology (which is a key issue in exposure models for seismic risk assessment). The results of the developed model achieved a precision of 93% and a recall of 95% when identifying nonductile buildings, which are the buildings most likely to be damaged in an earthquake. Identifying fine-grained material typology is more difficult, because many visual clues are physically hidden, but our model matches expert level performances, achieving a recall of 85% and accuracy scores ranging from 60% to 82% on the three most common building typologies, which account for 91% of the total building population in Medellín. Overall, this study shows that a CNN can make a substantial contribution to developing cost-effective exposure models. © 2020 Elsevier LtdÍtem Collaborative and distributed augmented reality in teaching multi-variate calculus(2006-01-01) Orozco, C.; Esteban, P.; Trefftz, H.; Orozco, C.; Esteban, P.; Trefftz, H.; Universidad EAFIT. Departamento de Ciencias; Educación Matemática e HistoriaThis article presents the first results of using an Augmented Reality (AR) tool, designed to support tutoring sessions in multi-variate calculus. The tool is used either in a face-to-face setting in which the instructor and the students are collocated or in a distance setting, in which the instructor and students are physically in remote places. The tool was used with two groups of students of Differential Calculus. The students had not been exposed to the concept of equations involving 3 variables and the corresponding surfaces in space. The experience explored how students generalized 2D graphics and equations with their 3D surfaces counterparts with the help of the tool.Ítem Collaborative virtual environments for teaching physics(SPRINGER, 2007-01-01) Giraldo, F.; Rojas, Á.M.J.; Esteban, P.; Trefftz, H.; Giraldo, F.; Rojas, Á.M.J.; Esteban, P.; Trefftz, H.; Universidad EAFIT. Departamento de Ciencias; Educación Matemática e HistoriaEafit University created a Telepresence application for distance education. It supports bidirectional audio and video and a shared virtual environment that allows the instructor and students to manipulate virtual objects in a collaborative manner, thereby reinforcing the learning process. In this paper we report the experience of creating new contents and deploying the tool in a Physics course at Universidad del Quindío. The tool was improved an extended in order to include new functionality and make it easier for development teams to create new content. Teaching for Understanding was used as pedagogical framework for course creation and also as a guide to develop the interviews that were used to capture user's perceptions and to asses the improvements in the learning process. The results showed that the use of the tool had a positive impact on the students' understanding of the topics at hand. © 2007 Springer.Ítem Design of pedagogic tools for teaching materials in product design engineering(The Design Society, 2015-01-01) Patiño Santa, L.F.The education in materials constitutes a vital part of the education of an engineer, since its comprehension determines the design and construction of products with appropriate materials that satisfy the current and future needs of the 21st century society [1]. In the Product Design Engineering degree, teaching and learning the attributes and applications of materials shouldn't be limited to lectures. This paper shows how through the design of pedagogic tools it is possible to go from a knowledge-transfer method to a student-centred method where motivation, teaching and learning strategies and problem-based learning, support meaningful learning [2] through the implementation of these tools in the classroom. © 2015, The Design Society. All rights reserved.Ítem Development of a model for evaluating the NPD process in SMEs: A Latin American experience(Elsevier, 2014-01-01) Hernandez, M.C.; Montoya, M.V.; Martínez, J.F.This article presents the findings resulting from a research project carried out in Medellin, Colombia, that was aimed at developing a model of diagnosis for the NPD process in manufacturing SMEs, adapted to the Latin American context. Four aspects were considered to develop the model: (i) Rational logic for problem solving based on the paradigm that a problem can be solved through systematic application of general principles and standardized knowledge, (ii) The Grounded Theory research approach, (iii) Integration of learning processes based on reflective and, (iv) Kolb's learning cycle as basic premises for the generation of a prescriptive model considering that the diagnosis process involves a learning process. Results obtained from the four companies selected for the study helped to validate an initial version. The model has proven to have great potential to improve the NPD process, as well as an opportunity to develop future comparative studies. Findings in the discussion and further reflection generated by the application of the four tools that encompasses the model within the companies, proved to be appropriated as a reflective practice and also a continuing learning process for SMEs. © 2014 Published by Elsevier B.V.Ítem Development of a model for evaluating the NPD process in SMEs: A Latin American experience(Elsevier, 2014-01-01) Hernandez, M.C.; Montoya, M.V.; Martínez, J.F.; Universidad EAFIT. Departamento de Administración; Estudios en Mantenimiento (GEMI)This article presents the findings resulting from a research project carried out in Medellin, Colombia, that was aimed at developing a model of diagnosis for the NPD process in manufacturing SMEs, adapted to the Latin American context. Four aspects were considered to develop the model: (i) Rational logic for problem solving based on the paradigm that a problem can be solved through systematic application of general principles and standardized knowledge, (ii) The Grounded Theory research approach, (iii) Integration of learning processes based on reflective and, (iv) Kolb's learning cycle as basic premises for the generation of a prescriptive model considering that the diagnosis process involves a learning process. Results obtained from the four companies selected for the study helped to validate an initial version. The model has proven to have great potential to improve the NPD process, as well as an opportunity to develop future comparative studies. Findings in the discussion and further reflection generated by the application of the four tools that encompasses the model within the companies, proved to be appropriated as a reflective practice and also a continuing learning process for SMEs. © 2014 Published by Elsevier B.V.Ítem A fractional Fourier transform-based method to detect impacts between the bogie and the car body of a railway vehicle: A data-driven approach(SAGE PUBLICATIONS LTD, 2018-01-01) Gutierrez-Carvajal, R. E.; Betancur, German R.; Castaneda, Leonel F.; Zajac, G.; Universidad EAFIT. Departamento de Ingeniería Mecánica; Estudios en Mantenimiento (GEMI)Structural railway transport elements are typically designed to work for at least 30 years without undergoing major maintenance. However, real-life operational conditions present behaviors different to the model predicted during the initial design phase, which affects the lifetime of the elements in question. This is the case of first-generation railway vehicles which operates in the city of Medellín, Colombia, as the bolster beam presented cracks after 12 years of operation, possibly due to undesired impacts between the bogie and the pivot of the bolster beam. Monitoring vibrational signals would give some sort of an insight into impact phenomena; however, herein lies the problem, as they are difficult to identify using only vibration signals, occurring during time events that take place in a speed-varying system. In this article, the authors present a technique that automatically detects impacts using multiple in-between time/frequency representations, ranking them according to their capacity to discriminate between impact events. Our results show that the best representation for this data was the Fractional Cepstrum Transform at order 0.5 (auROC = 0.961), which outperformed the best pure domain descriptor by least 4%. © 2016, © IMechE 2016.Ítem Modeling and control of nonlinear systems using an Adaptive LAMDA approach(Elsevier BV, 2020-01-01) Morales L.; Aguilar J.; Rosales A.; Chávez D.; Leica P.; Morales L.; Aguilar J.; Rosales A.; Chávez D.; Leica P.; Universidad EAFIT. Departamento de Ingeniería de Sistemas; I+D+I en Tecnologías de la Información y las ComunicacionesThis paper presents a soft computing technique for modeling and control of nonlinear systems using the online learning criteria. In order to obtain an accurate modeling, and therefore a controller with good performance, a method based on the fundamentals of the artificial intelligence algorithm, called LAMDA (Learning Algorithm for Multivariate Data Analysis), is proposed, with a modification of its structure and learning method that allows the creation of an adaptive approach. The novelty of this proposal is that for the first time LAMDA is used for fuzzy modeling and control of complex systems, which is a great advantage if the mathematical model is not available, partially known, or variable. The adaptive LAMDA consists of a training stage to establish initial parameters for the controller, and the application stage in which the control strategy is computed and updated using an online learning that evaluates the closed-loop system. We validate the method in several control tasks: (1) Regulation of mixing tank with variable dead-time (slow variable dynamics), (2) Regulation of a Heating, Ventilation and Air-Conditioning (HVAC) system (multivariable slow nonlinear dynamics), and (3) trajectory tracking of a mobile robot (multivariable fast nonlinear dynamics). The results of these experiments are analyzed and compared with other soft computing control techniques, demonstrating that the proposed method is able to perform an accurate control through the proposed learning technique. © 2020 Elsevier B.V.Ítem Probabilistic extension to the concurrent constraint factor oracle model for music improvisation(Asociacion Espanola de Inteligencia Artificial, 2016-01-01) Toro, M.; Toro, M.; Universidad EAFIT. Departamento de Ingeniería de Sistemas; I+D+I en Tecnologías de la Información y las ComunicacionesWe can program a Real-Time (RT) music improvisation system in C++ without a formal semantic or we can model it with process calculi such as the Non-deterministic Timed Concurrent Constraint (ntcc) calculus. “A Concurrent Constraints Factor Oracle (FO) model for Music Improvisation” (Ccfomi) is an improvisation model specified on ntcc. Since Ccfomi improvises non-deterministically, there is no control on choices and therefore little control over the sequence variation during the improvisation. To avoid this, we extended Ccfomi using the Probabilistic Non-deterministic Timed Concurrent Constraint calculus. Our extension to Ccfomi does not change the time and space complexity of building the FO, thus making our extension compatible with RT. However, there was not a ntcc interpreter capable of RT to execute Ccfomi. We developed Ntccrt -a RT capable interpreter for ntcc- and we executed Ccfomi on Ntccrt. In the future, we plan to extend Ntccrt to execute our extension to Ccfomi. © IBERAMIA and the authors.Ítem Teaching technical design specifications in product redesign processes using the FBS model(American Society of Mechanical Engineers (ASME), 2019-01-01) Montoya I.D.; Rendon-Velez E.; Gallego-Sanchez J.A.; Montoya I.D.; Rendon-Velez E.; Gallego-Sanchez J.A.; Universidad EAFIT. Departamento de Ingeniería Mecánica; Mecatrónica y Diseño de MáquinasEstablishment of technical design specifications for the redesign of technical systems, which are not derived from customers needs, relies heavily on designers intuition and experience. The very few methods proposed in the literature for the establishment of such technical design specifications hinders the learning process in conceptual design subjects. The objective of this article is to propose a structured method for identification of technical design specifications based on the function-behavior-state (FBS) model applied on redesign tasks. The method is being taught to students on the course of methodical design in order to develop their abilities on the identification of technical specifications without relying on previous knowledge on the system and relying more on detailed observation. This method is based on the observation and identification of the different states taken by the attributes of the entities conforming the system (i.e., flows-function carriers-environment). A small scale preliminary study was conducted in order to validate the initial performance of the proposed method. Eight groups made up of undergraduate mechanical engineering students with basic knowledge in conceptual design were instructed to identify technical design specifications. The redesign task was to create an automated solution for replacing the manual production process at small food manufacturing companies. Four groups of students had to establish the specifications using their experience while the other four groups had to use the proposed method. Initial results showed for the proposed method a 36% increase in the number of identified specifications. Copyright © 2019 ASME.Ítem Teaching technical design specifications in product redesign processes using the FBS model(American Society of Mechanical Engineers (ASME), 2019-01-01) Montoya I.D.; Rendon-Velez E.; Gallego-Sanchez J.A.Establishment of technical design specifications for the redesign of technical systems, which are not derived from customers needs, relies heavily on designers intuition and experience. The very few methods proposed in the literature for the establishment of such technical design specifications hinders the learning process in conceptual design subjects. The objective of this article is to propose a structured method for identification of technical design specifications based on the function-behavior-state (FBS) model applied on redesign tasks. The method is being taught to students on the course of methodical design in order to develop their abilities on the identification of technical specifications without relying on previous knowledge on the system and relying more on detailed observation. This method is based on the observation and identification of the different states taken by the attributes of the entities conforming the system (i.e., flows-function carriers-environment). A small scale preliminary study was conducted in order to validate the initial performance of the proposed method. Eight groups made up of undergraduate mechanical engineering students with basic knowledge in conceptual design were instructed to identify technical design specifications. The redesign task was to create an automated solution for replacing the manual production process at small food manufacturing companies. Four groups of students had to establish the specifications using their experience while the other four groups had to use the proposed method. Initial results showed for the proposed method a 36% increase in the number of identified specifications. Copyright © 2019 ASME.Ítem Technological tools to learn calculus(Institute of Electrical and Electronics Engineers Inc., 2015-12-02) Gómez, G.P.P.; Restrepo, C.M.Z.; Duarte, P.V.E.; Rivera, L.F.Z.; Gómez, G.P.P.; Restrepo, C.M.Z.; Duarte, P.V.E.; Rivera, L.F.Z.; Universidad EAFIT. Departamento de Ciencias; Educación Matemática e HistoriaOnline learning tools have allowed professors to carry out their classes in an interactive way, synchronic as well as asynchronic, giving them the opportunity to explore from different points of view specific themes or concepts, achieving greater dynamism in their classes through the active participation of students. This article presents the results of the implementation of an software tool for massive use designed with the objective of allowing Engineering students to strengthen their basic mathematical knowledge as well as to improve their results in courses such as Calculus I and Mathematics I. These courses are part of the first semesters of the syllabus for all undergraduate programs of the School of Engineering, and are basic courses in the formation of an engineer at Universidad EAFIT (Medellin, Colombia). This software tool for massive use allows students to self-diagnose, to solve exercises with different levels of complexity and difficulty, to visualize academic contents such as video classes and virtual resources, and to know their evolution in the understanding of basic concepts in calculus. On one hand, this facilitates the beginning of their studies at the university. On the other, it gives the professor an initial diagnose of the level students have to start the course so that continuous analytics can be performed based on the learning process of the student. Furthermore, this article shows the results of a comparative analysis done to two groups of students, a Control group and an Experimental group, that took Calculus I as part of their undergraduate studies. The experiment lasted two months with testing done at the beginning and at the end of the course. The objective was to register the level of knowledge acquired by the students and compare the differences between the two groups, control and experimental. The testing also allowed the progress of the student between tests to be measured, taking into account that the experimental group had the opportunity to explore the platform during this two-month period. Therefore, the analysis performed served to gather information useful for evaluating the effectiveness of the proposed system in the learning process of the students at the University. © 2015 IEEE.