Assessing Hwa-byung Vulnerability Using the Hwa-byung Personality Scale: a comparative study of machine learning...

Area of Science:

• Psychiatry and Behavioral Sciences
• Computational Psychology
• Health Informatics

Background:

• Hwa-byung (HB) is a culture-bound syndrome characterized by distress and somatic symptoms.
• Accurate identification of individuals vulnerable to HB is crucial for timely intervention.
• Existing assessment methods may benefit from enhanced predictive capabilities.

Purpose of the Study:

• To develop and compare machine learning models for classifying HB vulnerability.
• To evaluate the predictive efficacy of these models using an HB personality scale.
• To identify key personality predictors of HB symptom severity.

Main Methods:

• Data from 500 Korean adults (aged 19-44) were analyzed using HB personality and symptom scales.
• Machine learning models (Random Forest Classifier, XGBoost, Logistic Regression, and ensemble) were employed.
• Recursive feature elimination with cross-validation was used for feature selection and model evaluation.

Main Results:

• The 16-item HB personality scale served as optimal features for predicting HB vulnerability.
• An ensemble model (RFC-XGC-LR) achieved an accuracy of 0.80 and AUROC of 0.86 on the test set.
• Item 16 ('I often feel guilty easily') was the most significant predictor across all models.

Conclusions:

• Machine learning models, especially the ensemble method, show promise for screening individuals at risk for Hwa-byung based on personality traits.
• These models can improve the efficiency and accuracy of HB risk assessment in clinical settings.
• Early identification and intervention for HB can be facilitated through these data-driven approaches.