Criterios de información en la selección multimodelo de regresiones paramétricas: aplicaciones biológicas
Information criteria on multimodel selection of parametric regression: Biological applications
Additional Files
Abstract (en)
In carrying out modelling analysis of experimental data results important to obtain a measure of the relative fit of the models as a primary selection criterion. In this sense, there are few studies based on multi-model selection techniques for the theoretical representation of data sets, so it is common to incur in a misinterpretation of the existing patterns, or even more, the incorrect extrapolation and prediction based on the wrong model. This paper is intended to evaluate in 40 sets of data from various publications researches the effectiveness of the regression model designated by the authors by contrasting six regression models with the Akaike and Bayesian information criteria and to discuss its implications on subsequent interpretations made. It was found that the linear regression model was successful only in 13.35% of the datasets (AIC= 15%; BIC = 11.7%), but in the other hand, the logarithmic model was the most successful model in 38.5% of the cases (AIC= 35%; BIC= 41.1%) which casts doubt on the efficiency of the linear regression over other types of regression under biological data. It is clear then that the features discussed from regression analysis regardless multi-model selection depends on the subjectivity of the researcher and often incurs in selecting a model that involves greater losses of the information contained in the data set.Abstract (es)
Cuando se lleva a cabo análisis de modelamiento usando datos experimentales es importante obtener una medida de la confiabilidad del ajuste relativo de cada modelo como un criterio principal de seleccion. En este sentido, existen pocos estudios basados en técnicas de selección multimodelo para realizar representaciones teóricas de conjuntos de datos, por lo que es común incurrir en una mala interpretación de los patrones existentes, o más aún, extrapolar incorrectamente y basar predicciones en modelos equívocos. Este documentos está enfocado en evaluar en 40 conjuntos de datos provenientes de varios estudios ecológicos publicados, la efectividad de la regresión lineal designada por los autores al contrastarla con seis modelos de regresión usando los criterios de información de Akaike y Bayesiano, y además discutir las implicaciones de las interpretaciones subsecuentes de acuerdo al modelo. Se encontró que el modelo de regresión lineal fue exitoso en solo el 13.35% de los conjuntos de datos (15% de los conjuntos de datos para AIC y 11.7% de los datos para BIC), pero por otro lado, el modelo logarítmico fue m´as exitoso en el 38.5% de los casos (35% de los conjuntos de datos para AIC y 41.1% de los datos para BIC), generando dudas sobre la eficiencia del modelo de regresi´on lineal sobre los otros tipos de regresión en datos biológicos.
References
Akaike, H. (1998), “Information theory and an extension of the maximum likelihood principle”, in Selected Papers of Hirotugu Akaike, Springer, pp. 199–213.
Asp, M.L., Richardson, J.R., Collene, A.L., Droll, K.R. and Belury, M.A. (2012), “Dietary protein and beef consumption predict for markers of muscle mass and nutrition status in older adults”, The journal of nutrition, health & aging, Vol. 16 No. 9, pp. 784–790.
Bates, D.M. and Watts, D.G. (1988), Nonlinear regression: iterative estimation and linear approximations, Wiley Online Library.
Bevington, P.R. and Robinson, D.K. (1969), Data reduction and error analysis for the physical sciences,
How to Cite
License
The authors maintain the rights to the articles and therefore they are free to share, copy, distribute, execute and publicly communicate the work under the following conditions:
Recognize the credits of the work in the manner specified by the author or licensor (but not in a way that suggests that, you have their support or that they support your use of their work).
Comunicaciones en Estadística is licensed under Creative Commons Atribución-NoComercial-CompartirIgual 4.0 Internacional (CC BY-NC-SA 4.0)
Universidad Santo Tomás preserves the patrimonial rights (copyright) of the published works, and favors and allows the reuse of them under the aforementioned license.