2020 |
Umaquinga-Criollo, A. C.; Tamayo-Quintero, J. D.; Moreno-García, M. N.; Riascos, J. A.; Peluffo-Ordóñez, D. H. Multi-expert Methods Evaluation on Financial and Economic Data: Introducing Bag of Experts Journal Article In: Hybrid Artificial Intelligent Systems, no. 15th, pp. 437-449, 2020, ISBN: 978-3-030-61704-2. Abstract | BibTeX | Etiquetas: Bag of experts, clasificadores, Crowd-sourcing, Feed-forward neural network, Finance, Investment banking, MATLAB, multi-clasificadores, Multi-expert, patrones de desempeño académico, selección de características | Links: @article{umaquingame, The use of machine learning into economics scenarios results appealing since it allows for automatically testing economic models and predict consumer/client behavior to support decision-making processes. The finance market typically uses a set of expert labelers or Bureau credit scores given by governmental or private agencies such as Experian, Equifax, and Creditinfo, among others. This work focuses on introducing a so-named Bag of Expert (BoE): a novel approach for creating multi-expert Learning (MEL) frameworks aimed to emulate real experts labeling (human-given labels) using neural networks. The MEL systems “learn” to perform decision-making tasks by considering a uniform number of labels per sample or individuals along with respective descriptive variables. The BoE is created similarly to Generative Adversarial Network (GANs), but rather than using noise or perturbation by a generator, we trained a feed-forward neural network to randomize sampling data, and either add or decrease hidden neurons. Additionally, this paper aims to investigate the performance on economics-related datasets of several state-of-the-art MEL methods, such as GPC, GPC-PLAT, KAAR, MA-LFC, MA-DGRL, and MA-MAE. To do so, we develop an experimental framework composed of four tests: the first one using novice experts; the second with proficient experts; the third is a mix of novices, intermediate and proficient experts, and the last one uses crowd-sourcing. Our BoE method presents promising results and can be suitable as an alternative to properly assess the reliability of both MEL methods and conventional labeler generators (i.e., virtual expert labelers). |
Chamorro-Sangoquiza, Diana C.; Vargas-Muñoz, Andrés M.; Umaquinga-Criollo, Ana C.; Becerra, Miguel A.; Peluffo-Ordóñez, Diego H. Comparative study of data mining techniques to reveal patterns of academic performance in secondary education Journal Article In: RISTI - Revista Iberica de Sistemas e Tecnologias de Informacao, vol. E32, pp. 455–468, 2020, ISSN: 1646-9895. Abstract | BibTeX | Etiquetas: clasificadores, MATLAB, multi-clasificadores, patrones de desempeño académico, selección de características @article{chamorro2020_, Las técnicas de minería de datos permiten develar conocimiento a partir de grandes volúmenes de información, los cuales han sido poco exploradas en análisis de información de instituciones de educación, pero con una creciente demanda por este sector para apoyar la toma de decisiones. En esta investigación se propone una metodología de comparación de técnicas de minería de datos, aplicado al análisis de patrones académicos en estudiantes de educación media. Múltiples métodos de selección de atributos son aplicados para reducir la dimensionalidad y se comparan tres clasificadores y dos multi-clasificadores. Los experimentos se realizan en una base de datos de 285 instancias y 36 atributos obtenidos de una encuesta educativa aplicada a los alumnos del Tercer Año de Bachillerato de la Unidad Educativa Ibarra 2017-2018. Los mejores resultados de clasificación fueron alcanzados por los multiclasifiadores Boosted Tree y Bagged Tree con 93.24% de exactitud usando los datos seleccionados con el algoritmo BestFirst. |
2020 |
|
2. | A. C. Umaquinga-Criollo; J. D. Tamayo-Quintero; M. N. Moreno-García; J. A. Riascos; D. H. Peluffo-Ordóñez Multi-expert Methods Evaluation on Financial and Economic Data: Introducing Bag of Experts Journal Article In: Hybrid Artificial Intelligent Systems, no. 15th, pp. 437-449, 2020, ISBN: 978-3-030-61704-2. Abstract | Links | BibTeX | Etiquetas: Bag of experts, clasificadores, Crowd-sourcing, Feed-forward neural network, Finance, Investment banking, MATLAB, multi-clasificadores, Multi-expert, patrones de desempeño académico, selección de características @article{umaquingame, The use of machine learning into economics scenarios results appealing since it allows for automatically testing economic models and predict consumer/client behavior to support decision-making processes. The finance market typically uses a set of expert labelers or Bureau credit scores given by governmental or private agencies such as Experian, Equifax, and Creditinfo, among others. This work focuses on introducing a so-named Bag of Expert (BoE): a novel approach for creating multi-expert Learning (MEL) frameworks aimed to emulate real experts labeling (human-given labels) using neural networks. The MEL systems “learn” to perform decision-making tasks by considering a uniform number of labels per sample or individuals along with respective descriptive variables. The BoE is created similarly to Generative Adversarial Network (GANs), but rather than using noise or perturbation by a generator, we trained a feed-forward neural network to randomize sampling data, and either add or decrease hidden neurons. Additionally, this paper aims to investigate the performance on economics-related datasets of several state-of-the-art MEL methods, such as GPC, GPC-PLAT, KAAR, MA-LFC, MA-DGRL, and MA-MAE. To do so, we develop an experimental framework composed of four tests: the first one using novice experts; the second with proficient experts; the third is a mix of novices, intermediate and proficient experts, and the last one uses crowd-sourcing. Our BoE method presents promising results and can be suitable as an alternative to properly assess the reliability of both MEL methods and conventional labeler generators (i.e., virtual expert labelers). |
1. | Diana C. Chamorro-Sangoquiza; Andrés M. Vargas-Muñoz; Ana C. Umaquinga-Criollo; Miguel A. Becerra; Diego H. Peluffo-Ordóñez Comparative study of data mining techniques to reveal patterns of academic performance in secondary education Journal Article In: RISTI - Revista Iberica de Sistemas e Tecnologias de Informacao, vol. E32, pp. 455–468, 2020, ISSN: 1646-9895. Abstract | BibTeX | Etiquetas: clasificadores, MATLAB, multi-clasificadores, patrones de desempeño académico, selección de características @article{chamorro2020_, Las técnicas de minería de datos permiten develar conocimiento a partir de grandes volúmenes de información, los cuales han sido poco exploradas en análisis de información de instituciones de educación, pero con una creciente demanda por este sector para apoyar la toma de decisiones. En esta investigación se propone una metodología de comparación de técnicas de minería de datos, aplicado al análisis de patrones académicos en estudiantes de educación media. Múltiples métodos de selección de atributos son aplicados para reducir la dimensionalidad y se comparan tres clasificadores y dos multi-clasificadores. Los experimentos se realizan en una base de datos de 285 instancias y 36 atributos obtenidos de una encuesta educativa aplicada a los alumnos del Tercer Año de Bachillerato de la Unidad Educativa Ibarra 2017-2018. Los mejores resultados de clasificación fueron alcanzados por los multiclasifiadores Boosted Tree y Bagged Tree con 93.24% de exactitud usando los datos seleccionados con el algoritmo BestFirst. |