🎓 UpGrade

UpGrade helps students predict their exam scores based on current habits and, if desired, recommends habit adjustments to reach a target score, while keeping health and daily time constraints in mind.

🔍 Overview

• Predict Mode

  • Enter your current habits (study hours, sleep, diet, etc.)
  • Instantly see a predicted exam score via a trained linear regression model
  • Animated “count-up” percentage for a polished user experience

• Optimize Mode

  • Specify a desired exam score
  • UpGrade analyzes your habits and suggests which to tweak (sleep, study, diet, exercise, attendance, etc.)
  • Honors a 24 h daily budget and enforces healthy minimums (sleep ≥ 8 h, diet ≥ “Fair,” exercise ≥ once/week, mental health ≥ 6)
  • Color-coded, animated adjustments so you know exactly what to improve

🛠 Tech Stack

• Frontend:
  • Svelte (component-based UI)
  • Tailwind CSS (dark-mode theme + utility-style classes)
  • Svelte Motion (tweened) for smooth number animations
  • Fetch API to call backend endpoints

• Backend:
  • FastAPI (Python) with CORS enabled
  • scikit-learn LinearRegression model stored as model.joblib
  • joblib for serialization
  • Uvicorn as the ASGI server

• Data & Model Training:
  • Pandas & NumPy for data loading and preprocessing
  • A Jupyter notebook in research/ for initial exploration and training
  • train_model.py to retrain and dump the serialized model