Boston Housing Price Prediction

Michael Tamirat

Dataset Boston Housing Dataset from UCI Machine Learning Repository

506 samples with 14 features Target Variable: MEDV Primary Feature Used: RM

Data Cleaning & Feature Engineering

Data Loading: Loaded dataset from UCI repository with proper column names Feature Engineering: Created HOUSE_CONDITION feature from AGE column:

New: AGE ≤ 50 Old: AGE > 50 and ≤ 80 Very Old: AGE > 80

Feature Selection: Used RM (number of rooms) as predictor Data Split: 80% training, 20% testing

Model Performance Linear Regression Results:

R² Score: 0.3708 Mean Squared Error (MSE): 46.1448

Dependencies

pandas scikit-learn matplotlib numpy