KMAN Publication Inc. (KMANPUB) Applied Family Therapy Journal (AFTJ) 3041-8798 6 Serial Number 30 2025 11 01 Support Vector Machine Classification of Dysfunctional Family Systems Based on Psychological Assessment Data Support Vector Machine Classification of Dysfunctional Family Systems Based on Psychological Assessment Data 1 10 10.61838/kman.aftj.4950 EN Journal Article 2024 05 26 <p><strong>Objective:</strong> The objective of this study was to develop and validate a Support Vector Machine (SVM) model for the accurate classification of dysfunctional family systems using multidimensional psychological assessment data.</p> <p><strong>Methods and Materials: </strong>This quantitative cross-sectional study was conducted among 376 families in Armenia, including adolescents and their primary caregivers. Participants completed standardized psychological instruments assessing family functioning, parenting practices, parental stress, adolescent emotional and behavioral adjustment, and relational quality. A composite family dysfunction index was generated and used as the classification target. Data were preprocessed through normalization, imputation, and feature selection. The dataset was divided into training, validation, and test sets using stratified sampling. An SVM model with radial basis function kernel was trained and optimized through grid search and cross-validation. Model performance was evaluated using accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve. Feature contributions were interpreted using SHAP analysis.</p> <p><strong>Findings: </strong>The SVM model achieved an accuracy of 0.913, precision of 0.904, recall of 0.921, F1-score of 0.912, and AUC of 0.948 on the test dataset. Comparative analysis indicated that the SVM significantly outperformed logistic regression, k-nearest neighbors, and random forest classifiers (p &lt; .05). Feature importance analysis revealed that family functioning, parental stress, adolescent behavioral difficulties, parenting inconsistency, and marital adjustment were the strongest predictors of family dysfunction classification.</p> <p><strong>Conclusion: </strong>The findings demonstrate that SVM-based machine learning provides a highly accurate and robust framework for identifying dysfunctional family systems, offering substantial potential for early detection, targeted intervention, and data-driven mental health decision-making in family psychology and public health practice.</p> https://journals.kmanpub.com/index.php/aftj/article/download/4950/8910