Actualmente hablar de Inteligencia Artificial (IA) es para algunos hablar de robots y películas d... more Actualmente hablar de Inteligencia Artificial (IA) es para algunos hablar de robots y películas de ciencia ficción tales como, Terminator, Matrix, Yo Robot entre otras, para otros es aún un tema de reflexión. Si bien existe alguna base común, donde en algún punto divergen hacia caminos distintos para la creación de este tipo de películas, la Inteligencia Artificial se focaliza en otros objetivos, como por ejemplo Aprendizaje Automático y Aprendizaje Profundo, por lo que entenderlas, nos dará una visión clara hacia donde se dirige hoy la Inteligencia Artificial.
Climate change is here and is a reality in the world; therefore, studying this phenomenon based o... more Climate change is here and is a reality in the world; therefore, studying this phenomenon based on its relationship with meteorological parameters is the first step to making informed decisions. With this in mind, the objective of this work was to conduct a comparative analysis of machine learning techniques used in weather forecasting to evaluate their accuracy in weather forecasting in a localized area, Iquique. The methodology used was exploratory, and the design was experimen-tal based on Knowledge Discovery in Databases (KDD). The Transformer network and Arima in distant horizons gave better performance, indicating that Machine Learning techniques, particularly Deep Learning, can contribute to and complement classic weather forecasting techniques. Understanding the contribution of classic techniques such as Machine Learning in climate forecasting opens a range of possibilities to be further investigated.
Although the importance of Deep Learning has been well established in recent years, its role in c... more Although the importance of Deep Learning has been well established in recent years, its role in classifying objects in images is far from being understood in fine categories and this open problem remains to be solved in geometric shapes. Here we compare deep learning models using convolutional neural networks, in order to classify fine categories in geometrical figure type images. Through the proposed method we found that there are several configurations of base models that obtain accuracies close to 80%. The proposed method also allowed us to identify that using Transfer Learning increases the accuracy by about 7% compared to the base models. Overall, these data show that the number of examples plays an important role in obtaining good classification results, as well as their quality, since noisy data in a dataset can severely reduce the generalization performance of the model in question.
The objective of this work was to propose a methodology of data mining in civil engineering caree... more The objective of this work was to propose a methodology of data mining in civil engineering careers at the University of Tarapacá, Iquique headquarters. For which CRISP-DM was used. With this, it is possible to apply the different stages of the methodology to real data, depending on the problem to be solved, generating new knowledge and using the Rapidminer tool, with different measurements and algorithms. Then, it is concluded that this methodology can generate new knowledge based on established steps, having the possibility of applying innovative practices in data management.
2020 39th International Conference of the Chilean Computer Science Society (SCCC), 2020
Computer-related careers have maintained the stigma of being mostly masculine. Currently, such ca... more Computer-related careers have maintained the stigma of being mostly masculine. Currently, such careers are demanded in labour market, but there are not enough professionals to meet demand and less than 20% of the students enrolled in technology-related careers are women, according to the Chilean Higher Education Information Services. The absence of information that characterizes the women who enroll in the Computer related area in Chile is the main motivation for this work.This study presents a comparison of the results of classification techniques for the data set of female students who choose Computer Science in universities belonging to the Council of Rectors of Chilean Universities (CRUCH), in order to identify relevant variables to choose this carrers. School location, academic performance, and mother's education were relevant. The results of two resampling schemes for imbalanced classes are similar, however Naiive Bayes with undersampling obtained slightly more balanced results with Prediction of 61%.
Revista Ibérica de Sistemas y Tecnologías de Información, 2020
En el Hospital Regional Ernesto Torres Galdames, de la ciudad de Iquique, en la región de Tarapac... more En el Hospital Regional Ernesto Torres Galdames, de la ciudad de Iquique, en la región de Tarapacá-Chile, se implementó una solución en base al uso de herramientas de Inteligencia de Negocios, para el apoyo en la toma de decisiones asociada a la información de prácticas profesionales. Se consideró un análisis de los procesos de la Unidad de Docencia, para conocer su modelo de procesos, realizar la selección de las herramientas a implementar, establecer los procesos de Extracción, Transformación y Carga, y la presentación de los datos y reportes. En base a la Metodología de Kimball se construyó un data mart, cuyo uso permitió a la Unidad de Docencia observar tendencias, cantidades de estudiantes registrados y, analizar y coordinar de mejor forma los cupos de prácticas profesionales.
El adulto mayor se ha alejado de las tecnologías de la información creyendo no contar con habilid... more El adulto mayor se ha alejado de las tecnologías de la información creyendo no contar con habilidades para usarlas en su vida cotidiana, particularmente lo relacionado con procesos cognitivos. Por lo anterior, el objetivo de este trabajo fue medir el grado de satisfacción de los Adultos Mayores usando una WebApp de estimulación cognitiva. La investigación fue de tipo exploratorio, donde se investigaron procesos cognitivos y patrones de diseño de interfaces de software, para usarlos en el desarrollo de la aplicación web y construir ejercicios de estimulación cognitiva. Finalmente se buscó conocer si los patrones de diseño facilitaron la utilización de esta aplicación mediante una encuesta a 30 personas en 3 clubes de adultos mayores, para medir el grado de satisfacción. Entre los resultados se evidenció que los patrones de diseño de interfaces de gráficas para adultos mayores facilitan el uso y comprensión de las funcionalidades del software, brindándoles la posibilidad de usar un computador y desarrollar habilidades cognitivas. Se concluye que la aplicación logró su objetivo, puesto que la mayoría de estas personas no tuvieron problemas con las funcionalidades de la aplicación, motivando a futuros trabajos en desarrollo de conductas de autocuidado, interacciones sociales y aspectos psicológicos del adulto mayor.
The objective of this work was to propose a methodology of data mining in civil engineering caree... more The objective of this work was to propose a methodology of data mining in civil engineering careers at the University of Tarapacá, Iquique headquarters. For which CRISP-DM was used. With this, it is possible to apply the different stages of the methodology to real data, depending on the problem to be solved, generating new knowledge and using the Rapidminer tool, with different measurements and algorithms. Then, it is concluded that this methodology can generate new knowledge based on established steps, having the possibility of applying innovative practices in data management.
Academic performance is a complex multifactorial phenomenon that depends on the educational origi... more Academic performance is a complex multifactorial phenomenon that depends on the educational origin. From here, it is analyzed the influence of educational background on the academic performance of 215 civil engineering students at the University of Tarapacá. The methodology is based on a quantitative study of multiple linear regression, with cross-sectional data. The predictive capacity of the University Selection Test is questioned, whose conclusions show that the score obtained is not a good predictor of academic performance, since it does not directly influence, unlike school dependence, which has a higher predictive value.
The objective of this work was to design and execute a technological tool that supports the manag... more The objective of this work was to design and execute a technological tool that supports the management and planning of the activities carried out in the classrooms by the teacher, through a stage of research, development of a decision tree and finally tests and analysis of results. With this, a comparison of the results of the use of the software versus the empirical knowledge of the teacher was made, achieving high levels of effectiveness of the use of the decision tree technique for the knowledge and modeling of the student, in order to favor the diagnosis of the course. Finally, Artificial Intelligence techniques can be an important support for the understanding of teaching- learning processes.
RESUMEN Se desarrolló una herramienta de tipo WEB para apoyar a los profesionales de la salud enc... more RESUMEN Se desarrolló una herramienta de tipo WEB para apoyar a los profesionales de la salud encargados de asignar interconsultas médicas a pacientes con riesgo de cáncer mamario, ocupando árboles de decisión como método para clasificar si el paciente cumple o no los requisitos para que su solicitud sea aceptada según criterios establecidos. Esta clasificación permite generar una lista de pacientes que necesitan una atención médica oportuna, información que es recopilada por el sistema para desplegar una lista ordenada según el resultado de una mamografía y la fecha de la solicitud de interconsulta. Conociendo esta lista, el profesional encargado podrá enviarla al Servicio de Orientación Médico Estadístico (SOME), entidad encargada de asignar hora y fecha a cada solicitud de interconsulta, cuya respuesta también podrá ser almacenada en el sistema. Conclusión: Mediante una demostración y posterior encuesta, el proyecto se califica como un aporte al proceso de asignación de interconsultas médicas a pacientes con riesgo de cáncer mamario. PALABRAS CLAVES: Cáncer de mama, sistema Web, inteligencia artificial, árboles de decisión SUMMARY A Web tool which is described was developed in order to give support to health care professionals which assign consultations to people in risk of breast cancer. They use the decision tree as the method to classify patients and also to determine if they meet the requirements in order to approve their application according to the criteria established by health care professionals. This classification allows generating a list of patients who need medical assistance as soon as possible. This information is gathered by the system in order to display an organized list according to the mammogram results and the date of the consultation in the application form. Once the list is done, the person in charge will be able to send the list to "Servicio de Orientación Médico Estadístico", an organization who assigns the date of consultations for each application form and also stores the answer in the system if needed. Conclusion: Through a demonstration and subsequent survey, the project qualifies as a contribution to the process of assigning medical interconsultations patients with breast cancer risk.
Actualmente hablar de Inteligencia Artificial (IA) es para algunos hablar de robots y películas d... more Actualmente hablar de Inteligencia Artificial (IA) es para algunos hablar de robots y películas de ciencia ficción tales como, Terminator, Matrix, Yo Robot entre otras, para otros es aún un tema de reflexión. Si bien existe alguna base común, donde en algún punto divergen hacia caminos distintos para la creación de este tipo de películas, la Inteligencia Artificial se focaliza en otros objetivos, como por ejemplo Aprendizaje Automático y Aprendizaje Profundo, por lo que entenderlas, nos dará una visión clara hacia donde se dirige hoy la Inteligencia Artificial.
Climate change is here and is a reality in the world; therefore, studying this phenomenon based o... more Climate change is here and is a reality in the world; therefore, studying this phenomenon based on its relationship with meteorological parameters is the first step to making informed decisions. With this in mind, the objective of this work was to conduct a comparative analysis of machine learning techniques used in weather forecasting to evaluate their accuracy in weather forecasting in a localized area, Iquique. The methodology used was exploratory, and the design was experimen-tal based on Knowledge Discovery in Databases (KDD). The Transformer network and Arima in distant horizons gave better performance, indicating that Machine Learning techniques, particularly Deep Learning, can contribute to and complement classic weather forecasting techniques. Understanding the contribution of classic techniques such as Machine Learning in climate forecasting opens a range of possibilities to be further investigated.
Although the importance of Deep Learning has been well established in recent years, its role in c... more Although the importance of Deep Learning has been well established in recent years, its role in classifying objects in images is far from being understood in fine categories and this open problem remains to be solved in geometric shapes. Here we compare deep learning models using convolutional neural networks, in order to classify fine categories in geometrical figure type images. Through the proposed method we found that there are several configurations of base models that obtain accuracies close to 80%. The proposed method also allowed us to identify that using Transfer Learning increases the accuracy by about 7% compared to the base models. Overall, these data show that the number of examples plays an important role in obtaining good classification results, as well as their quality, since noisy data in a dataset can severely reduce the generalization performance of the model in question.
The objective of this work was to propose a methodology of data mining in civil engineering caree... more The objective of this work was to propose a methodology of data mining in civil engineering careers at the University of Tarapacá, Iquique headquarters. For which CRISP-DM was used. With this, it is possible to apply the different stages of the methodology to real data, depending on the problem to be solved, generating new knowledge and using the Rapidminer tool, with different measurements and algorithms. Then, it is concluded that this methodology can generate new knowledge based on established steps, having the possibility of applying innovative practices in data management.
2020 39th International Conference of the Chilean Computer Science Society (SCCC), 2020
Computer-related careers have maintained the stigma of being mostly masculine. Currently, such ca... more Computer-related careers have maintained the stigma of being mostly masculine. Currently, such careers are demanded in labour market, but there are not enough professionals to meet demand and less than 20% of the students enrolled in technology-related careers are women, according to the Chilean Higher Education Information Services. The absence of information that characterizes the women who enroll in the Computer related area in Chile is the main motivation for this work.This study presents a comparison of the results of classification techniques for the data set of female students who choose Computer Science in universities belonging to the Council of Rectors of Chilean Universities (CRUCH), in order to identify relevant variables to choose this carrers. School location, academic performance, and mother's education were relevant. The results of two resampling schemes for imbalanced classes are similar, however Naiive Bayes with undersampling obtained slightly more balanced results with Prediction of 61%.
Revista Ibérica de Sistemas y Tecnologías de Información, 2020
En el Hospital Regional Ernesto Torres Galdames, de la ciudad de Iquique, en la región de Tarapac... more En el Hospital Regional Ernesto Torres Galdames, de la ciudad de Iquique, en la región de Tarapacá-Chile, se implementó una solución en base al uso de herramientas de Inteligencia de Negocios, para el apoyo en la toma de decisiones asociada a la información de prácticas profesionales. Se consideró un análisis de los procesos de la Unidad de Docencia, para conocer su modelo de procesos, realizar la selección de las herramientas a implementar, establecer los procesos de Extracción, Transformación y Carga, y la presentación de los datos y reportes. En base a la Metodología de Kimball se construyó un data mart, cuyo uso permitió a la Unidad de Docencia observar tendencias, cantidades de estudiantes registrados y, analizar y coordinar de mejor forma los cupos de prácticas profesionales.
El adulto mayor se ha alejado de las tecnologías de la información creyendo no contar con habilid... more El adulto mayor se ha alejado de las tecnologías de la información creyendo no contar con habilidades para usarlas en su vida cotidiana, particularmente lo relacionado con procesos cognitivos. Por lo anterior, el objetivo de este trabajo fue medir el grado de satisfacción de los Adultos Mayores usando una WebApp de estimulación cognitiva. La investigación fue de tipo exploratorio, donde se investigaron procesos cognitivos y patrones de diseño de interfaces de software, para usarlos en el desarrollo de la aplicación web y construir ejercicios de estimulación cognitiva. Finalmente se buscó conocer si los patrones de diseño facilitaron la utilización de esta aplicación mediante una encuesta a 30 personas en 3 clubes de adultos mayores, para medir el grado de satisfacción. Entre los resultados se evidenció que los patrones de diseño de interfaces de gráficas para adultos mayores facilitan el uso y comprensión de las funcionalidades del software, brindándoles la posibilidad de usar un computador y desarrollar habilidades cognitivas. Se concluye que la aplicación logró su objetivo, puesto que la mayoría de estas personas no tuvieron problemas con las funcionalidades de la aplicación, motivando a futuros trabajos en desarrollo de conductas de autocuidado, interacciones sociales y aspectos psicológicos del adulto mayor.
The objective of this work was to propose a methodology of data mining in civil engineering caree... more The objective of this work was to propose a methodology of data mining in civil engineering careers at the University of Tarapacá, Iquique headquarters. For which CRISP-DM was used. With this, it is possible to apply the different stages of the methodology to real data, depending on the problem to be solved, generating new knowledge and using the Rapidminer tool, with different measurements and algorithms. Then, it is concluded that this methodology can generate new knowledge based on established steps, having the possibility of applying innovative practices in data management.
Academic performance is a complex multifactorial phenomenon that depends on the educational origi... more Academic performance is a complex multifactorial phenomenon that depends on the educational origin. From here, it is analyzed the influence of educational background on the academic performance of 215 civil engineering students at the University of Tarapacá. The methodology is based on a quantitative study of multiple linear regression, with cross-sectional data. The predictive capacity of the University Selection Test is questioned, whose conclusions show that the score obtained is not a good predictor of academic performance, since it does not directly influence, unlike school dependence, which has a higher predictive value.
The objective of this work was to design and execute a technological tool that supports the manag... more The objective of this work was to design and execute a technological tool that supports the management and planning of the activities carried out in the classrooms by the teacher, through a stage of research, development of a decision tree and finally tests and analysis of results. With this, a comparison of the results of the use of the software versus the empirical knowledge of the teacher was made, achieving high levels of effectiveness of the use of the decision tree technique for the knowledge and modeling of the student, in order to favor the diagnosis of the course. Finally, Artificial Intelligence techniques can be an important support for the understanding of teaching- learning processes.
RESUMEN Se desarrolló una herramienta de tipo WEB para apoyar a los profesionales de la salud enc... more RESUMEN Se desarrolló una herramienta de tipo WEB para apoyar a los profesionales de la salud encargados de asignar interconsultas médicas a pacientes con riesgo de cáncer mamario, ocupando árboles de decisión como método para clasificar si el paciente cumple o no los requisitos para que su solicitud sea aceptada según criterios establecidos. Esta clasificación permite generar una lista de pacientes que necesitan una atención médica oportuna, información que es recopilada por el sistema para desplegar una lista ordenada según el resultado de una mamografía y la fecha de la solicitud de interconsulta. Conociendo esta lista, el profesional encargado podrá enviarla al Servicio de Orientación Médico Estadístico (SOME), entidad encargada de asignar hora y fecha a cada solicitud de interconsulta, cuya respuesta también podrá ser almacenada en el sistema. Conclusión: Mediante una demostración y posterior encuesta, el proyecto se califica como un aporte al proceso de asignación de interconsultas médicas a pacientes con riesgo de cáncer mamario. PALABRAS CLAVES: Cáncer de mama, sistema Web, inteligencia artificial, árboles de decisión SUMMARY A Web tool which is described was developed in order to give support to health care professionals which assign consultations to people in risk of breast cancer. They use the decision tree as the method to classify patients and also to determine if they meet the requirements in order to approve their application according to the criteria established by health care professionals. This classification allows generating a list of patients who need medical assistance as soon as possible. This information is gathered by the system in order to display an organized list according to the mammogram results and the date of the consultation in the application form. Once the list is done, the person in charge will be able to send the list to "Servicio de Orientación Médico Estadístico", an organization who assigns the date of consultations for each application form and also stores the answer in the system if needed. Conclusion: Through a demonstration and subsequent survey, the project qualifies as a contribution to the process of assigning medical interconsultations patients with breast cancer risk.
Uploads
Papers by Jorge Diaz-Ramirez