Predicting Health Insurance Costs Using Machine Learning

In this project, we build a machine learning model to predict health insurance costs based on customer demographics and lifestyle factors. The project takes you step by step through the entire data science workflow — from data cleaning and exploration to building, evaluating, and deploying predictive models. We achieve an R² score of 84%, demonstrating strong predictive performance.

Additionally, we deploy the model as an interactive web application using Streamlit, allowing users to input features and see real-time predictions for health insurance costs.

---

Live Demo

You can access the live deployed application here: Health Insurance Cost Predictor App

Dashboard Preview

Here’s a screenshot of the interactive Streamlit dashboard:

Table of Contents

• Project Overview
• Dataset
• Technologies Used
• Installation
• Usage
• Modeling
• Evaluation
• Deployment
• License

---

Project Overview

The goal of this project is to predict an individual’s health insurance cost based on features such as:

• Age
• Sex
• Body Mass Index (BMI)
• Number of children
• Smoking status
• Region

We follow a complete machine learning workflow:

1. Data Cleaning & Preprocessing: Handle missing values, encode categorical variables, scale features.
2. Exploratory Data Analysis (EDA): Visualize patterns and relationships between features and target variable.
3. Modeling: Train multiple regression models including:
   • Linear Regression
   • Polynomial Regression
   • Random Forest Regressor
   • Support Vector Regressor (SVR)
   • XGBoost Regressor
4. Model Evaluation: Compare models using R², RMSE, and MAE metrics.
5. Deployment: Build a Streamlit app for interactive predictions.

---

Dataset

The dataset contains the following columns:

Column	Description
age	Age of the individual
sex	Gender (male/female)
bmi	Body Mass Index
children	Number of children covered by health insurance
smoker	Smoking status (yes/no)
region	Residential region (northeast, northwest, southeast, southwest)
charges	Health insurance charges (target variable)

You can find the dataset in Kaggle: Medical Cost Personal Dataset.

---

Technologies Used

• Python 3.9+
• Pandas, NumPy
• Matplotlib, Seaborn
• Scikit-learn
• XGBoost
• Streamlit

---

Installation

1. Clone this repository:

git clone https://github.com/Dharmendra-07/Health-Insurance-Cost-Predictor.git
cd health-insurance-predictor

2. Create a virtual environment (optional but recommended):

python3 -m venv venv
source venv/bin/activate   # On Mac/Linux
venv\Scripts\activate      # On Windows

3. Install required packages:

pip3 install -r requirements.txt

---

Usage

1. Train Models

Run the Jupyter notebook to explore data, train models, and evaluate their performance:

jupyter notebook

2. Run Streamlit App

Start the interactive web app:

python3 -m streamlit run streamlit_app.py

Open the displayed URL in your browser to interact with the predictor.

---

Modeling

We experimented with multiple regression algorithms:

Model	R² Score
Linear Regression	0.75
Polynomial Regression	0.79
Random Forest Regressor	0.82
SVR	0.78
XGBoost Regressor	0.84

The XGBoost Regressor achieved the best performance and is used in the Streamlit app.

---

Evaluation

We evaluated models using:

• R² Score (coefficient of determination)
• Root Mean Squared Error (RMSE)
• Mean Absolute Error (MAE)

Visualizations and detailed evaluation metrics can be found in the notebook.

---

Deployment

The model is deployed as an interactive Streamlit app. Features:

• Input user data (age, BMI, smoking status, etc.)
• Predict health insurance cost in real-time
• User-friendly interface with sliders and dropdowns

Run the app:

python3 -m streamlit run streamlit_app.py

---

License

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