NeuroSleep Insight is a Streamlit-based sleep quality and performance prediction system developed as part of an undergraduate Computer Science final-year research project.
The system uses machine learning to predict sleep quality from lifestyle and health-related factors, then provides performance and neuroscience-inspired interpretation based on sleep, fatigue, attention, productivity, and brain-network function.
Sleep Pattern Analysis and Performance Prediction
- Predicts Quality of Sleep from user input
- Provides Performance Readiness interpretation
- Shows Fatigue Risk, Focus Level, and Productivity Insight
- Includes a Neuroscience Insight section
- Connects prediction results to concepts such as:
- functional connectivity
- brain-state stability
- memory consolidation
- attention networks
- sleep-related performance
The machine learning model predicts Quality of Sleep using selected lifestyle and health-related features.
- Sleep Duration
- Stress Level
- Physical Activity Level
- Daily Steps
- Heart Rate
- Age
- Gender
- BMI Category
- Sleep Disorder
- Occupation
| Model | MAE | MSE | RMSE | R² Score |
|---|---|---|---|---|
| Decision Tree Regressor | 0.039 | 0.021 | 0.145 | 0.986 |
| Random Forest Regressor | 0.045 | 0.026 | 0.160 | 0.983 |
| Linear Regression | 0.121 | 0.051 | 0.227 | 0.966 |
The Decision Tree Regressor was selected as the best-performing model and saved as:
sleep_quality_model.pkl