Prodigy Info Tech - House Price Prediction Using Linear Regression

As part of my virtual Machine Learning Internship I have completed the task of predicting house price using linear regression.

Goal

Predicting house price based on:

• Square Feet (GrLivArea)
• Number of bedrooms (BedroomAbvGr)
• Number of bathrooms (FullBath)

Implementation Details

As per task requirements, I used:

1. Linear Regression model
2. 3 independent features:
   • GrLivArea (Square feet)
   • BedroomAbvGr (Number of bedrooms)
   • FullBath (Number of bathrooms)

Model Performance Metrics

Metric	Value
Mean Absolute Error	35788.0612924363
Mean Squared Error	2806426667.247853
Root Mean Squared Error	52975.71771338122
R² Score	0.6341189942328371

This indicates that the model performance is decent, with average score.

Reasons for Descent Score

• Used only 3 features
• If multiple features were implemented, linear regression might not be the best choice
• Possible overfitting to the independent features, especially since target values are in dollars

Results

1. Saved the trained model in pickle format
2. Implemented a Python script that allows users to predict house prices based on their input

-Mithun