CADOES

CADOES estimates biological sex from pelvic osteometric measurements using machine-learning classification algorithms that leverage pelvic sexual dimorphism for forensic and bioarchaeological sex assessment.

Key Features:

• Population-specific models: Uses the J.A. Serra (1938) Coimbra Identified Skeletal Collection comprising 256 individuals (131 females, 125 males) and 38 metric variables to build osteometric models.
• Machine-learning classification algorithms: Applies various classification algorithms to osteometric data, reporting accuracies of 85%–92% with three variables and 85.33%–97.33% with all 38 variables.
• Osteometric modeling: Constructs sex-estimation models based on pelvic metric variables.

Scientific Applications:

• Probabilistic prediction: Produces probabilistic estimates of skeletal sex for use in forensic anthropology and bioarchaeology.
• Educational utility: Illustrates application of osteometry and statistical classification in human skeletal sex estimation for teaching and research training.

Methodology:

Uses osteometric measurements from the J.A. Serra (1938) Coimbra Identified Skeletal Collection (256 individuals, 38 metric variables) and applies machine-learning classification algorithms to build sex-estimation models.