KMAN Publication Inc. (KMANPUB) Psychology of Woman Journal 3041-8615 7 Serial Number 27 2026 05 01 CatBoost-Based Prediction of Marital Satisfaction in Women From Attachment Security, Emotional Regulation, Relationship Commitment, and Communication Patterns CatBoost-Based Prediction of Marital Satisfaction in Women From Attachment Security, Emotional Regulation, Relationship Commitment, and Communication Patterns 1 13 10.61838/kman.pwj.5459 EN Journal Article 2026 01 07 <table> <tbody> <tr> <td> <p><strong>Objective:</strong> The present study aimed to develop and evaluate a CatBoost-based machine learning model for predicting marital satisfaction among married women using attachment security, emotional regulation, relationship commitment, and communication patterns as predictor variables.</p> <p><strong>Methods and Materials:</strong> This cross-sectional predictive study was conducted among 742 married women residing in Taiwan. Participants were recruited through community organizations, social networks, and online survey platforms. Data were collected using the Revised Dyadic Adjustment Scale (RDAS) to assess marital satisfaction, the Experiences in Close Relationships-Revised Questionnaire (ECR-R) to measure attachment security, the Emotion Regulation Questionnaire (ERQ) to evaluate emotional regulation, the Commitment Inventory to assess relationship commitment, and the Communication Patterns Questionnaire-Short Form (CPQ-SF) to examine communication patterns. Following data preprocessing and screening procedures, the dataset was divided into training (80%) and testing (20%) subsets. A CatBoost regression algorithm was developed using optimized hyperparameters identified through five-fold cross-validation. Model performance was evaluated using the coefficient of determination (R²), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). Feature importance and SHapley Additive exPlanations (SHAP) analyses were performed to determine the relative contribution and interpretability of each predictor variable.</p> <p><strong>Findings:</strong> The CatBoost model demonstrated excellent predictive performance, explaining 84.7% of the variance in marital satisfaction within the testing dataset (R² = 0.847). Relationship commitment emerged as the most influential predictor (35.84%), followed by communication patterns (28.63%), attachment security (21.57%), and emotional regulation (13.96%). Correlation analyses revealed significant positive associations between marital satisfaction and attachment security (r = .68, p &lt; .01), emotional regulation (r = .59, p &lt; .01), relationship commitment (r = .76, p &lt; .01), and communication patterns (r = .72, p &lt; .01). SHAP analyses further indicated that higher levels of all predictor variables contributed positively to marital satisfaction predictions, while lower levels reduced predicted outcomes.</p> </td> </tr> </tbody> </table> <p><strong>Conclusion:</strong> The findings demonstrate that marital satisfaction among women can be predicted with high accuracy using a CatBoost machine learning model. Relationship commitment, communication patterns, attachment security, and emotional regulation represent critical determinants of marital satisfaction, with commitment emerging as the strongest predictor. The study highlights the value of integrating relationship psychology with advanced machine learning techniques to improve understanding, prediction, and intervention planning for marital well-being.</p> https://www.journals.kmanpub.com/index.php/psywoman/article/download/5459/10733