Placement Prediction Web Application

A FastAPI-based machine learning web app that predicts a student's placement likelihood based on soft skills scores.

---

Live Demo

🔗 https://placement-softv2.onrender.com
(If not loaded instantly, Render may take 30-60 seconds to wake up)

---

Application Features

✅ User-friendly UI (Bootstrap + Jinja Template)
✅ Real-time Placement Prediction
✅ Confidence Percentage Display
✅ Full REST API Support
✅ Model trained using Voting Classifier (RF + SVC + GB)
✅ Automatic /docs API documentation by FastAPI
✅ Deployed and Production Ready

---

Model Inputs (Expected Score Range)

Feature	Description	Range
Mock HR Interview Score	Soft skill HR score	5 – 20
Group Discussion Score	Communication & teamwork	5 – 20
Presentation Score	Clarity + speaking skills	5 – 20
English CEFR Level	(1=A1 → 6=C2)	1 – 6
English Score	English test score	5 – 20

Additional Features created automatically inside backend:

• HR_GD_interaction = MockHR × GD
• English_Total = English CEFR × English Score
• Overall_Performance = mean(MockHR, GD, Presentation)

---

ML Model

Technique	Used?
RandomForest	✅
GradientBoosting	✅
Support Vector Classifier	✅
SMOTE Oversampling	✅
Standard Scaling	✅
GridSearch Hyperparameter Tuning	✅

Final model saved as: model/p3.pkl

---