Boston House Price Prediction using Linear Regression – A Beginner ML Project

🔍 Overview

In this project, I built a Machine Learning model to predict house prices using the Boston Housing Dataset. The model uses Linear Regression, one of the most fundamental algorithms in supervised learning, to predict prices based on multiple housing features.

---

🧠 Problem Statement

The goal is to predict the median value of owner-occupied homes (MEDV) in Boston suburbs using features like crime rate, average number of rooms per dwelling, property tax rate, etc.

---

📦 Dataset Used

• Dataset: Boston Housing Dataset (csv file)

• Features: 13 variables including crime rate, room count, property tax, etc.

• Target: MEDV (Median value of homes in $1000's)

---

🔧 Tools & Libraries

• Python

• Pandas, NumPy – Data handling

• Matplotlib, Seaborn – Visualization

• Scikit-learn – ML modeling and evaluation

---

📊 Exploratory Data Analysis (EDA)

I began with basic EDA to understand:

• Feature distributions

• Missing values (if any)

• Correlation between features and target

📌 Correlation Heatmap revealed that:

• RM (average number of rooms per dwelling) is positively correlated with MEDV

• LSTAT (percentage of lower status population) is negatively correlated

---

🧪 Model Building

1. Feature selection

2. Train-test split (80-20)

3. Linear Regression model training

4. Prediction & Evaluation

---

📈 Evaluation Metrics

• R² Score: 0.999998

• Mean Absolute Error (MAE): 0.0091

These results indicate a very high accuracy and a minimal error, showing the model fits the dataset exceptionally well.

---

📷 Visualizations

• Correlation Matrix

• Actual vs Predicted Prices plot

• Residuals analysis

You can embed your plots/screenshots here if available

---

📚 Key Learnings

• How to handle real-world structured data

• Importance of correlation and feature scaling

• Building and evaluating a regression model

• Measuring model performance using R² and MAE

---

🔗 Project Demo

• Video Demo: [https://www.linkedin.com/posts/lokeshpatidar-n15_machinelearning-linearregression-mlprojects-activity-7314188458715979776-M45N?utm_source=share&utm_medium=member_desktop&rcm=ACoAAEhLJ2EBgrbFlNdm77-JVglYmofpxUBumXg]