Individual-specific postural discomfort prediction using decision tree models.

Area of Science:

• Ergonomics and Human Factors
• Computational Modeling
• Biomedical Engineering

Background:

• Understanding individual posture-discomfort relationships is crucial for personalized ergonomics.
• Existing models often lack the granularity to capture inter-individual variability.
• Decision tree (DT) algorithms offer a flexible approach for complex data analysis.

Purpose of the Study:

• To explore the application of the decision tree (DT) algorithm for individual-level modeling of posture-discomfort relationships.
• To assess the interpretability and accuracy of DT models in predicting postural discomfort.
• To investigate inter-individual variability in discomfort perception using DT modeling.

Main Methods:

• Generated individual-specific DT models using postural discomfort rating data.
• Utilized seated upper body postures, postural variables, and task variables as inputs.
• Employed DT algorithm for its robustness with non-linear, noisy data and interpretability.

Main Results:

• Individual-specific DT models accurately predicted postural discomfort.
• Significant inter-individual variability was observed in the DT modeling outcomes.
• The models demonstrated the utility of DT for analyzing complex human perception data.

Conclusions:

• DT modeling is a valuable tool for understanding individual posture-discomfort relationships.
• The approach facilitates the development of personalized digital human models.
• DT modeling enhances the investigation of human discomfort perception processes and inter-individual differences.