Combining Bayesian Calibration and Copula Models for Age Estimation.

Area of Science:

• Forensic Anthropology
• Biometrics
• Statistical Modeling

Background:

• Accurate chronological age estimation is critical for undocumented individuals, particularly forced migrants.
• Existing methods often rely on single predictors (e.g., dental/skeletal maturity) or independent ordinal predictors.
• There is a need for more robust and accurate age prediction methods using multiple continuous anthropometric indicators.

Purpose of the Study:

• To develop and evaluate a novel method for chronological age prediction using multiple continuous anthropometric indices.
• To combine Bayesian calibration with copula statistical modeling for enhanced age estimation.
• To compare the accuracy of the proposed method against existing single-predictor and independent models.

Main Methods:

• Implementation of a copula-based statistical model to integrate multiple continuous anthropometric predictors.
• Utilized Bayesian calibration as a robust framework for age prediction.
• Comparative analysis of the copula-based method against independent and single-predictor models.

Main Results:

• The proposed copula-based method demonstrated increased accuracy in chronological age prediction.
• Combining multiple continuous indices via copulas significantly improved predictive performance.
• The novel approach outperformed traditional single-predictor models.

Conclusions:

• The copula-based Bayesian calibration method offers a more accurate approach to chronological age estimation from anthropometric data.
• This method is particularly valuable in contexts where identity documents are unavailable, such as with forced migrants.
• Future research can explore further integration of diverse anthropometric variables for even greater precision.