A High-Precision SVM Model for Depression Detection Among Farmers

Abstract

Depression among farmers is an increasing concern, driven by a complex interplay of agricultural, socioeconomic, and behavioral stressors. This study employs machine learning to identify depression in farmers by integrating diverse factors such as economic conditions, social behavior, and farming-related challenges. This study offers a novel data-driven framework for the detection of depression among Indian farmers by merging socioeconomic and psychological indicators—an area with limited prior investigation. The integration of Patient Health Questionnaire 9 (PHQ-9) assessments with an SVM-RFE model establishes a robust and accurate method for large-scale mental health screening in agrarian populations. Data were gathered from 1,069 farmers in Mandya, India, using structured surveys and the validated PHQ-9. The SVM model demonstrated high performance, achieving 96.59% accuracy, 96.47% precision, 96.59% recall, and a 96.00% F1-score, surpassing several other advanced classification algorithms. The findings underscore the significant impact of economic instability, social isolation, and limited access to mental healthcare on farmers’ psychological well-being. Notably, gender-based disparities emerged, with approximately 40% of female farmers found to be more susceptible to depression. The study highlights the urgent need for integrated mental health support systems and agricultural policy reforms, advocating for scalable, AI-powered early detection tools tailored to rural farming communities.