A Hybrid Approach Variable Selection Algorithm Based on Mutual Information for Data-Driven Industrial Soft-Sensor Applications

Un algoritmo de selección de variables de enfoque híbrido basado en información mutua para aplicaciones de sensores blandos industriales basados en datos

dc.creatorCote-Ballesteros, Jorge E.
dc.creatorGrisales Palacios, Victor Hugo
dc.creatorRodriguez-Castellanos, Jhon Edisson
dc.date2022-06-03
dc.date2023-03-22T18:49:14Z
dc.date2023-03-22T18:49:14Z
dc.date.accessioned2023-09-06T17:59:11Z
dc.date.available2023-09-06T17:59:11Z
dc.identifierhttps://revistas.unimilitar.edu.co/index.php/rcin/article/view/5644
dc.identifier10.18359/rcin.5644
dc.identifierhttp://hdl.handle.net/10654/42617
dc.identifier.urihttps://repositorioslatinoamericanos.uchile.cl/handle/2250/8693723
dc.descriptionThe development of virtual sensors predicting the desired output requires a careful selection of input variables for model construction. In an industrial environment, datasets contain many instrumentation system measures; however, these variables are often non-relevant or excessive information. This paper proposes a variable selection algorithm based on mutual information examination, redundancy analysis, and variable reduction for soft-sensor modeling. A relevance calculation is performed in the first stage to select important variables using the mutual information criterion. Then, the detection and exclusion of redundant variables are carried out, penalizing undesired variables. Finally, the most relevant variables subset is determined through a wrapper method using Mallowssans' Cp metric to assess the fitting prediction performance. The approach was successfully applied to estimate the ethanol concentration for a distillation column process using an adaptive network-based fuzzy inference system architecture as a non-linear dynamic regression model. A comparative study was performed considering the application of correlation analysis and the method proposed in this study. Simulation results show the effectiveness of the proposed approach in the variable selection providing a reduction in search of suitable models that achieve faster results for developing soft sensors oriented to industrial applications.
dc.descriptionEl desarrollo de sensores virtuales que predicen el resultado o producto deseado requie- re una cuidadosa selección de variables de entrada para la construcción del modelo. En un entorno industrial, los conjuntos de datos contienen muchas medidas del sistema de instrumentación; sin embargo, estas variables suelen ser información no relevante o excesiva. Este artículo propone un algoritmo de selección de variables basado en el examen de información mutua, el análisis de re- dundancia y la reducción de variables para el modelado de sensores blandos. En la primera etapa se realiza un cálculo de relevancia para seleccionar variables importantes utilizando el criterio de infor- mación mutua. Luego, se realiza la detección y exclusión de variables redundantes, penalizando las variables no deseadas. Finalmente, el subconjunto de variables más relevante se determina a través de un método de envoltura utilizando la métrica Cp de Mallows para evaluar el rendimiento de la pre- dicción de ajuste. El enfoque se aplicó con éxito para estimar la concentración de etanol para un pro- ceso de columna de destilación utilizando una arquitectura de sistema de inferencia difusa basada en red adaptativa como un modelo de regresión dinámica no lineal. Se realizó un estudio comparativo considerando la aplicación del análisis de correlación y el método propuesto en este estudio. Los re- sultados de la simulación muestran la efectividad del enfoque propuesto en la selección de variables proporcionando una reducción en la búsqueda de modelos adecuados que logren resultados más rápidos para el desarrollo de sensores blandos orientados a 
dc.formatapplication/pdf
dc.formattext/xml
dc.languageeng
dc.publisherUniversidad Militar Nueva Granada
dc.relationhttps://revistas.unimilitar.edu.co/index.php/rcin/article/view/5644/5112
dc.relationhttps://revistas.unimilitar.edu.co/index.php/rcin/article/view/5644/5185
dc.relation/*ref*/B. Lin, B. Recke, J. K. H. Knudsen, and S. B. Jørgensen, "A systematic approach for soft sensor development," Comput. Chem. Eng., vol. 31, no. 5-6, pp. 419-425, 2007. doi: https://doi.org/10.1016/j.compchemeng.2006.05.030
dc.relation/*ref*/P. Kadlec, B. Gabrys, and S. Strandt, "Data-driven Soft Sensors in the process industry," Computers and Chemical Engineering, vol. 33, no. 4. pp. 795-814, 2009. doi: https://doi.org/10.1016/j.compchemeng.2008.12.012
dc.relation/*ref*/I. Guyon, A. Elisseeff, and A. M. De, "An Introduction to Variable and Feature Selection," J. Mach. Learn. Res., vol. 3, pp. 1157-1182, 2003.
dc.relation/*ref*/R. Kohavi and G. H. John, "Wrappers for feature subset selection," Artif. Intell., vol. 97, no. 1-2, pp. 273-324, 1997. doi: https://doi.org/10.1016/S0004-3702(97)00043-X
dc.relation/*ref*/S. Visalakshi and V. Radha, "A literature review of feature selection techniques and applications: Review of feature selection in data mining," 2014 IEEE Int. Conf. Comput. Intell. Comput. Res. IEEE ICCIC 2014, no. 1997, 2015. doi: https://doi.org/10.1109/ICCIC.2014.7238499
dc.relation/*ref*/L. Fortuna, S. Graziani, and M. G. Xibilia, "Soft sensors for product quality monitoring in debutanizer distillation columns," Control Eng. Pract., vol. 13, no. 4, pp. 499-508, 2005. doi: https://doi.org/10.1016/j.conengprac.2004.04.013
dc.relation/*ref*/S. B. Chitralekha and S. L. Shah, "Application of support vector regression for developing soft sensors for nonlinear processes," Can. J. Chem. Eng., vol. 88, no. 5, pp. 696-709, 2010. doi: https://doi.org/10.1002/cjce.20363
dc.relation/*ref*/E. Y. Nagai, L. Valeria, and R. De Arruda, "Soft sensor based on Fuzzy Model indentification."
dc.relation/*ref*/M. Liukkonen, E. Hälikkä, T. Hiltunen, and Y. Hiltunen, "Dynamic soft sensors for NOx emissions in a circulating fluidized bed boiler," Appl. Energy, vol. 97, no. x, pp. 483-490, 2012. doi: https://doi.org/10.1016/j.apenergy.2012.01.074
dc.relation/*ref*/A. Rogina, I. Šiško, I. Mohler, Z. ̌ Ujević, and N. Bolf, "Soft sensor for continuous product quality estimation (in crude distillation unit)," Chem. Eng. Res. Des., vol. 89, no. January, pp. 2070-2077, 2011. doi: https://doi.org/10.1016/j.cherd.2011.01.003
dc.relation/*ref*/X. Yuan, H. Yang, and N. S. Wang, "A method of variables selection for soft sensor based on distributed mutual information," vol. 7, no. 3, pp. 1164-1169, 2015.
dc.relation/*ref*/F. Souza, R. Araújo, S. Soares, and J. Mendes, "VARIABLE SELECTION BASED ON MUTUAL INFORMATION FOR SOFT SENSORS APPLICATIONS," in Proceedings of the 9th Portuguese Conference on Automatic Control (Controlo 2010), 2009.
dc.relation/*ref*/Q. Li, X. Du, W. Liu, and W. Ba, "Soft sensor modelling based on mutual information variable selection and partial least squares," Proc. - 2017 Chinese Autom. Congr. CAC 2017, vol. 2017-Janua, pp. 3649-3654, 2017. doi: https://doi.org/10.1109/CAC.2017.8243414
dc.relation/*ref*/F. Curreri, S. Graziani, and M. G. Xibilia, "Input selection methods for data-driven Soft sensors design: Application to an industrial process," Inf. Sci. (Ny)., 2020. doi: https://doi.org/10.1016/j.ins.2020.05.028
dc.relation/*ref*/F. Curreri, G. Fiumara, and M. G. Xibilia, "Input selection methods for soft sensor design: A survey," Futur. Internet, vol. 12, no. 6, pp. 1-24, 2020. doi: https://doi.org/10.3390/fi12060097
dc.relation/*ref*/V. H. Alves Ribeiro and G. Reynoso-Meza, "Feature selection and regularization of interpretable soft sensors using evolutionary multi-objective optimization design procedures," Chemom. Intell. Lab. Syst., vol. 212, no. February, p. 104278, 2021. doi: https://doi.org/10.1016/j.chemolab.2021.104278
dc.relation/*ref*/T. M. Cover and J. A. Thomas, Elements of Information Theory, Second Edi. Wiley Jhon & sons, 2006.
dc.relation/*ref*/C. L. Mallows, "Some Comments on Cp," Technometrics, vol. 15, no. November, pp. 87-94, 1973. doi: https://doi.org/10.2307/1267380
dc.relation/*ref*/R. Battiti, "Using Mutual Information for Selecting Features in Supervised Neural-Net Learning," Ieee Trans. Neural Networks, vol. 5, no. 4, pp. 537-550, 1994. doi: https://doi.org/10.1109/72.298224
dc.relation/*ref*/H. Peng, F. Long, and C. Ding, "Feature selection based on mutual information: Criteria of Max-Dependency, Max-Relevance, and Min-Redundancy," IEEE Trans. Pattern Anal. Mach. Intell., vol. 27, no. 8, pp. 1226-1238, 2005. doi: https://doi.org/10.1109/TPAMI.2005.159
dc.relation/*ref*/L. Gao and W. Wu, "Relevance assignation feature selection method based on mutual information for machine learning," Knowledge-Based Syst., vol. 209, p. 106439, Dec. 2020. doi: https://doi.org/10.1016/j.knosys.2020.106439
dc.relation/*ref*/M. Mittal, S. C. Satapathy, V. Pal, B. Agarwal, L. M. Goyal, and P. Parwekar, "Prediction of coefficient of consolidation in soil using machine learning techniques," Microprocess. Microsyst., vol. 82, p. 103830, Apr. 2021. doi: https://doi.org/10.1016/j.micpro.2021.103830
dc.relation/*ref*/D. Effrosynidis and A. Arampatzis, "An evaluation of feature selection methods for environmental data," Ecol. Inform., vol. 61, no. January, p. 101224, 2021. doi: https://doi.org/10.1016/j.ecoinf.2021.101224
dc.relation/*ref*/J. S. R. Jang, "ANFIS: Adaptive-Network-Based Fuzzy Inference System," IEEE Trans. Syst. Man Cybern., vol. 23, no. 3, pp. 665-685, 1993. doi: https://doi.org/10.1109/21.256541
dc.relation/*ref*/J.-S. R. Jang, "Input selection for ANFIS learning," Proc. IEEE 5th Int. Fuzzy Syst., vol. 2, pp. 1493-1499, 1996.
dc.rightsDerechos de autor 2022 Ciencia e Ingeniería Neogranadina
dc.sourceCiencia e Ingenieria Neogranadina; Vol. 32 No. 1 (2022); 59-70
dc.sourceCiencia e Ingeniería Neogranadina; Vol. 32 Núm. 1 (2022); 59-70
dc.sourceCiencia e Ingeniería Neogranadina; v. 32 n. 1 (2022); 59-70
dc.source1909-7735
dc.source0124-8170
dc.subjectsoft-sensor
dc.subjectfeature selection
dc.subjectmutual information
dc.subjectindustrial processes
dc.subjectdata-driven
dc.subjectdistillation column
dc.subjectSensores virtuales
dc.subjectselección de características
dc.subjectinformación mutua
dc.subjectprocesos industriales
dc.subjectbasado en datos
dc.subjectcolumna de destilación
dc.titleA Hybrid Approach Variable Selection Algorithm Based on Mutual Information for Data-Driven Industrial Soft-Sensor Applications
dc.titleUn algoritmo de selección de variables de enfoque híbrido basado en información mutua para aplicaciones de sensores blandos industriales basados en datos
dc.typeinfo:eu-repo/semantics/article
dc.typeinfo:eu-repo/semantics/publishedVersion


Este ítem pertenece a la siguiente institución