Stress Indicators Dataset for Mental Health Classification
Description
This dataset captures responses from 2,000 students regarding their experiences with stress, health, academics, and emotional well-being. The original dataset was collected via Google Forms survey and later expanded to ensure sufficient sample size for machine learning research. The survey items used a five-point Likert scale (1 = Not at all to 5 = Extremely). All responses have been anonymized to protect participants’ privacy. The dataset is suitable for: Stress detection and classification (Eustress vs. Distress). Analyzing correlations between psychological, academic, and social stressors. Developing machine learning models for student mental health monitoring. 🔑 Key Features 👤 Demographics gender: 0 = Male, 1 = Female age: Numeric (18–22) 🧠 Emotional & Stress Indicators stress_experience heartbeat_palpitations anxiety_tension sleep_problems restlessness irritability sadness_low_mood loneliness_isolation concentration_problems 🩺 Physical & Health Indicators headaches health_issues weight_changes 📚 Academic & Environment Stressors academic_overload peer_competition low_academic_confidence subject_confidence academic_conflicts class_attendance professor_difficulties work_environment home_environment 💬 Social & Relationship Factors relationship_stress lack_relaxation_time 🎯 Target Variable stress_type (encoded): 0 = Distress (Negative Stress) 1 = Eustress (Positive Stress) 2 = Other / Mixed Stress 📂 Dataset File File: Stress_Dataset_Polished_Encoded.csv Rows: 2,000 Columns: 26 Missing Values: None Duplicates: None Target Variable: stress_type (3 classes)
Files
Institutions
- Daffodil International University