Insurance-Premium-Category

Insurance Premium Category

🛡️ InsureIQ — Insurance Premium Category Predictor

A complete end-to-end Machine Learning project that predicts a user's insurance premium category (Low / Medium / High) based on personal, health, and lifestyle inputs.

Built with Scikit-learn, served via FastAPI, and visualized through a polished Streamlit UI.

---

📌 Table of Contents

• [Overview]
• [Project Structure]
• [How It Works]
• [Features]
• [Tech Stack]
• [Installation & Setup]
  • [Run Streamlit UI]
  • [Run FastAPI Backend]
• [API Usage]
• [Dataset]
• [Model Details]
• [Screenshots]
• [License]

---

📖 Overview

InsureIQ takes a user's demographic and lifestyle inputs and classifies them into one of three insurance premium brackets:

Category	Meaning
🟢 Low	Low-risk profile — affordable premiums likely
🟡 Medium	Moderate risk — standard plans recommended
🔴 High	Elevated risk — comprehensive coverage advised

The project exposes two interfaces:

• A FastAPI REST endpoint (/predict) for programmatic access
• A Streamlit web app for interactive, browser-based predictions

---

📁 Project Structure

InsureIQ/
│
├── app.py                          # FastAPI backend with /predict endpoint
├── streamlit_app.py                # Streamlit frontend UI
├── ML_Model_used_in_FastAPI.ipynb  # Jupyter notebook — EDA, training, evaluation
├── model.pkl                       # Trained & serialized ML model (pickle)
├── insurance.csv                   # Raw dataset used for training
├── requirement.txt                 # Python dependencies
└── README.md                       # You're here!

---

⚙️ How It Works

Input Fields

Field	Type	Description
age	int	Age of the user (1–119)
weight	float	Weight in kilograms
height	float	Height in meters
income_lpa	float	Annual income in Lakhs Per Annum (₹)
smoker	bool	Whether the user smokes
city	str	City of residence
occupation	str	One of 10 supported occupations

Derived Features (computed automatically)

These are calculated from the raw inputs and fed into the ML model:

Feature	Formula / Logic
bmi	weight / height²
age_group	young / adult / middle_aged / senior
lifestyle_risk	high / medium based on smoker + BMI
city_tier	1 (metro) / 2 (tier-2) / 3 (other)

---

✨ Features

• ✅ Real-time BMI calculator with color-coded category display
• ✅ Clean Streamlit UI with dark theme, live inputs, and styled result cards
• ✅ FastAPI backend with Pydantic validation and auto-generated Swagger docs
• ✅ Derived feature computation mirrors both frontend and backend — results are consistent
• ✅ City classification into Tier 1 / 2 / 3 automatically
• ✅ Model trained on real-world-style insurance data with 10 occupations and 50+ cities

---

🧰 Tech Stack

Layer	Technology
ML Model	Scikit-learn
Model Serialization	Pickle
Backend API	FastAPI + Uvicorn
Data Validation	Pydantic v2
Frontend UI	Streamlit
Data Processing	Pandas, NumPy
Notebook	Jupyter

---

🚀 Installation & Setup

Prerequisites

• Python 3.9+
• pip

1. Clone the Repository

git clone https://github.com/your-username/InsureIQ.git
cd InsureIQ

2. Install Dependencies

pip install -r requirement.txt

Note: If you face any version conflicts, create a virtual environment first:

python -m venv venv
source venv/bin/activate        # macOS/Linux
venv\Scripts\activate           # Windows
pip install -r requirement.txt

---

Run Streamlit UI

streamlit run streamlit_app.py

Then open your browser at: http://localhost:8501

---

Run FastAPI Backend

uvicorn app:app --reload

Then open your browser at: http://127.0.0.1:8000/docs

⚠️ The app only has a POST /predict endpoint — use the /docs Swagger UI to test it, not the root URL directly.

---

🔌 API Usage

Endpoint

POST /predict

Request Body (JSON)

{
  "age": 35,
  "weight": 75.5,
  "height": 1.75,
  "income_lpa": 12.5,
  "smoker": false,
  "city": "Mumbai",
  "occupation": "Engineer"
}

Response

{
  "predicted_category": "low"
}

Example with curl

curl -X POST "http://127.0.0.1:8000/predict" \
  -H "Content-Type: application/json" \
  -d '{
    "age": 45,
    "weight": 90,
    "height": 1.70,
    "income_lpa": 8.0,
    "smoker": true,
    "city": "Delhi",
    "occupation": "Driver"
  }'

Example with Python requests

import requests

payload = {
    "age": 45,
    "weight": 90,
    "height": 1.70,
    "income_lpa": 8.0,
    "smoker": True,
    "city": "Delhi",
    "occupation": "Driver"
}

response = requests.post("http://127.0.0.1:8000/predict", json=payload)
print(response.json())
# Output: {'predicted_category': 'high'}

---

📊 Dataset

File: insurance.csv

Column	Type	Description
age	int	Age of the individual
weight	float	Body weight (kg)
height	float	Height (meters)
income_lpa	float	Annual income in LPA
smoker	str	yes / no
city	str	City of residence
occupation	str	Profession category
insurance_premium_category	str	Target — low / medium / high

---

🤖 Model Details

The ML pipeline and feature engineering are documented in ML_Model_used_in_FastAPI.ipynb.

Key steps in the notebook:

1. Exploratory Data Analysis (EDA)
2. Feature Engineering — BMI, age group, lifestyle risk, city tier
3. Encoding categorical variables
4. Model training & hyperparameter tuning
5. Evaluation (accuracy, classification report)
6. Serialization to model.pkl via pickle

Supported Occupations: Engineer, Business, Doctor, Farmer, Driver, Teacher, Private Job, Student, Government Job, Freelancer

City Tiers:

• Tier 1: Mumbai, Delhi, Bangalore, Chennai, Hyderabad, Pune
• Tier 2: Jaipur, Lucknow, Bhopal, Nagpur, Indore, and 40+ more
• Tier 3: All other cities

---

📄 License

This project is open-source and available under the MIT License.

---

Built with ❤️ using Python, FastAPI, and Streamlit.