Use of Bayesian Model Averaging to Estimate Model Uncertainty for Predicting Strain in a Four-Layered Flexible Pavement

Rema, Aswathy; Swamy, Aravind Krishna

doi:10.1061/ajrua6.0001123

Cited by 2 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This result is consistent with the results obtained since the length of the slab is the most critical variable in the co-kriging model. It could be observed in-situ that the dimensions of the slabs do not comply with the AASHTO (1993) recommendations [14], as the lengths are almost three times the typical measurement of 4.5 m and the widths are twice the 3.6 m measurement. The use of the kriging geostatistical method for data estimation was beneficial in predicting the number of cracks in the pavement under study; applied to the engineered area, it is enough of a prediction method, provided the necessary data is available.…”

Section: Resultsmentioning

confidence: 90%

“…A more recent study has also used Bayesian modeling to identify the most influential variables and to quantify the uncertainty of linear regression models predicting the occurrence of deformation in flexible pavement structures [14]. It was possible to establish the variables with the most significant variability and those with the greatest uncertainty by using the best models developed in conjunction with the Bayesian technique.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Statistical Study Based on the Kriging Method and Geographic Mapping in Rigid Pavement Defects in Southern Chile

2022

View full text Add to dashboard Cite

ASTM D6433 is used to assess the need for maintenance of pavement sections. Although the Pavement Condition Index (PCI) factor calculation method provides reliable values, this method analyzes sections and defects individually and indicates current maintenance needs, but it cannot be used to predict the occurrence of new defects. Therefore, it is necessary to complement this method by considering variables that influence the occurrence of faults, among which are the geospatial distribution and the specific characteristics of the slabs. This research focuses on the identification of multiple types of disturbances that exist in Portland Cement Pavements (PCC), located in a high traffic area in the city of Valdivia (Chile). A spatial geostatistical relationship is established through visual inspection using geographical maps, as well as distribution, using the kriging method. This technique makes use of variograms that allow quantifying the parameters used in this study, thus expressing the spatial autocorrelation of the faults analyzed. From the results obtained by spatial geostatistics and kriging, it is possible to generate a data correlation for the distribution and characteristics of the streets considered. In addition, a co-kriging method is established instead of an ordinary kriging method. The relationship between observed and predicted values improved from 0.3327 to 0.5770. The width of the slabs, as well as some streets, is shown in our analysis to be unimportant. For better model accuracy, the number of covariates associated with the type of vehicle traffic, the age and shape of the slabs, and the construction techniques used for the pavement needs to increase.

show abstract

Section: Resultsmentioning

confidence: 90%

Section: Introductionmentioning

confidence: 99%

Statistical Study Based on the Kriging Method and Geographic Mapping in Rigid Pavement Defects in Southern Chile

2022

View full text Add to dashboard Cite

show abstract

“…The maximuml formula ( ( `|e)) in equation ( 7) is the model in f, which has a high posterior probability value. The posterior probability of each significant model parameter is determined by averaging the posterior probabilities of each parameter from the selected best models [18]. If a model has a value of X Y greater than d = 20 , the model will be eliminated.…”

Section: Bayesian Model Averaging (Bma)mentioning

confidence: 99%

“…The Bayesian Model Averaging (BMA) approach offers the best alternative solution to this problem [8], [9]. Bayesian model averaging has been applied in other fields such as economic and financial [10]- [13], hydrology [14]- [17], engineering [18], behavioral research [19], environmental [20], agriculture [21], climate and meteorological [22]- [24] and many others.…”

Section: Introductionmentioning

confidence: 99%

Bayesian Model Averaging (BMA) Based on Logistic Regression for Gene Selection and Classification of Animal Tumor Disease on Microarray Data

Kuswanto¹,

Fitriana²

2022

Int. J. Adv. Sci. Eng. Inf. Technol.

View full text Add to dashboard Cite

Tumor is one of the deadly diseases which is frequently to be found in animals. However, identifying whether an animal has a tumor still becomes a big challenge. Classification of tumor disease can be done through gene expression, which consists of hundreds of genes, but only a small number of samples is taken. This data structure is called microarray data having the characteristic of highdimensional data. The choice of a single model can be a problem for high-dimensional data because it ignores model uncertainty. This research proposed to use Bayesian Model Averaging (BMA) to model the uncertainty model by averaging the posterior distribution of all best models, weighted by their posterior model probabilities. Selecting relevant genes to diagnose animal tumors is very important; hence, variable selection needs to be carried out. The selection of predictor variables is carried out by using the iterative BMA algorithm.The BMA results showed that from 335 gene expressions, 12 genes were selected to be relevant genes for classifying whether the animals have a tumor or normal. Moreover, from 2 335 possible models formed, 12 of the best models are selected. The accuracy of BMA results is assessed using the Brier Score, resulting from a value indicating that the BMA model is good enough to classify animals, whether they have a tumor or not. This research has proven that BMA with logistic performance has very good predictability; hence, the method can be applied to classify other diseases.

show abstract

Use of Bayesian Model Averaging to Estimate Model Uncertainty for Predicting Strain in a Four-Layered Flexible Pavement

Cited by 2 publications

References 28 publications

Statistical Study Based on the Kriging Method and Geographic Mapping in Rigid Pavement Defects in Southern Chile

Statistical Study Based on the Kriging Method and Geographic Mapping in Rigid Pavement Defects in Southern Chile

Bayesian Model Averaging (BMA) Based on Logistic Regression for Gene Selection and Classification of Animal Tumor Disease on Microarray Data

Contact Info

Product

Resources

About