Breastfeeding and Caries. A Relationship Analysis to Develop a Computer-Assisted Diagnosis Using Random Forest

摘要

This work proposes the analysis of a set of features related to breastfeeding and dental caries. The data was taken from NHANES 2013-2014, for the development of the model, the random forest algorithm is implemented, was evaluated through a statistical analysis based on the out-of-bag error (OOB), the receiver operating characteristic (ROC) curve and the area under the curve (AUC). Results; AUC = 0.669 and OOB = 36% were observed, which shows that the multivari-ate model developed is statistically significant and achieves an adequate generalization for the classification of the subjects. Conclusions; A model based on information related to breastfeeding is able to classify subjects with presence of dental caries in a generalized way, obtaining preliminary results for the development of a support tool for the diagnosis of this condition based on a new approach.