A scroll-stopping, student-first Machine Learning project that turns mess feedback into actionable predictions. 🧠✨
Ever stood outside the mess thinking: “Is today worth it?” 😭🍛
This project uses Linear Regression to help students estimate a mess meal’s expected rating based on measurable factors like:
- 🍲 Food Quality
- 🧼 Cleanliness
- 📦 Quantity
- 👅 Taste
By training a regression model on historical feedback, the tool learns patterns between these inputs and the final rating. Students can then use the model to make smarter decisions—so the next trip to the mess isn’t a gamble. 🎯📈
| Feature | What it does | Why it matters |
|---|---|---|
| ⭐ Rating Prediction | Predicts expected mess rating | Helps students decide fast |
| 📊 Feedback Analytics | Analyzes trends in feedback | Makes quality issues visible |
| 🤖 ML-ready Workflow | Train → evaluate → save model | Easy for contributors |
| 🎚️ Simple Linear Model | Uses Linear Regression | Fast, interpretable baseline |
| Category | Tools |
|---|---|
| Language | Python 🐍 |
| Data Handling | Pandas 🧮 |
| Machine Learning | Scikit-Learn 🤖 |
| Model | Linear Regression 📉 |
pip install -r requirements.txtIf
requirements.txtisn’t present yet, install manually:
pip install pandas scikit-learn matplotlibpython model/train_model.pyThis trains a Linear Regression model and saves it as a model.pkl file.
python analysis.pyGenerates visualization for feedback analysis.
- 📅 Predictive Weekly Menu: Recommend the best days and predicted meal quality for the week.
- 🤖 Telegram Alert Bot: Notify students when predicted rating crosses a threshold (e.g., ⭐ 4.2+). 📲
- 📝 Sentiment Analysis on text feedback: Convert comments/reviews into sentiment scores and combine them with numeric metrics.
- 🧾 Auto data collection pipeline: Stream feedback from forms/spreadsheets into the dataset automatically.
- 📈 Better models & evaluation: Compare Linear Regression with trees/ensembles and show interpretable results.
- 🌐 Live web dashboard: Host the predictor + analytics in a lightweight UI.
Welcome, GSSoC'26 participants! 🙌
This repo is intentionally built for learning and fast iteration. If you’re new to ML, you can still contribute by:
- Improve the README and project documentation 🧾
- Add EDA plots and data cleaning steps 📊
- Write tests for training/analysis scripts 🧪
- Enhance visualization for insights 📈
- Implement feature engineering (e.g., scaling, interactions) 🧠
- Add cross-validation and better evaluation metrics 📏
- Build a simple UI for predictions (Streamlit/Flask) 🌐
- Build a Telegram bot for alerts 🤖📲
- Add sentiment analysis combining text + numeric features 📝
- Upgrade the model pipeline (MLflow, pipelines, model registry) 🏗️
How to start:
- Fork the repository 🍴
- Pick a roadmap item you like ✅
- Create a branch:
blackboxai/<your-name>-<feature> - Submit a PR with clear explanation ✍️
data/→ dataset (mess_data.csv)model/→ training scripts + saved modelanalysis.py→ analytics and visualizationapp.py→ app layer (if used)
Linear Regression is a strong baseline because it is:
- ✅ Simple & interpretable
- ✅ Quick to train
- ✅ Great for establishing early impact
As the project grows, future work can compare it with non-linear models while keeping interpretability as a priority. 🔍
Add your chosen license details here (e.g., MIT, Apache-2.0). 🧷
Built for students, by students. 💙🎓
If you’re using this project as part of GSSoC'26, drop a star 🌟—it helps students find and use it faster!