Comparison of binary classifiers in forensic dentistry for sex determination

In forensic odontology, teeth are essential components of the stomatognathic system, since they constitute excellent material for research, thanks to their almost indestructible structure and therefore the information linked to their size and characteristics, is extremely useful in determining sex and age. In this study, which corresponds to a sample of 524 lower plaster models (286 male subjects and 238 females) of patients assisted in an orthodontic clinic in Montevideo, Uruguay, the mesiodistal diameter (width), and gingivo-incisal height of the 2 canines and the intercanine distance are measured for sex identification based on odontometric measurements and their ratios. Different binary classifiers from what is known as supervised classification methods are applied in the statistical learning paradigm. Several methods are proposed and their performance is evaluated by working with test and learning samples and evaluating different metrics to measure the accuracy of the models, showing a performance greater than 65%. The details of this project are on the OSF platform in the project. Relaciones Odontométricas en Odontología Forense (Odontometric Relationships in Forensic Dentestry) en https://osf.io/javru/