Predicting-Housing-Prices-using-Machine-Leaning

Overview

This repository contains two machine learning projects that predict housing prices using different techniques:

1. Linear Regression applied to housing data from Dumbo, Brooklyn.
2. XGBoost Regression applied to housing data from Iowa.

Datasets

• The Dumbo, Brooklyn dataset includes various property attributes and market factors affecting house prices in an urban setting.
• The Iowa dataset provides structured real estate data from a suburban/rural environment, including categorical and numerical features.

Methodologies

1. Linear Regression (Dumbo, Brooklyn)

• Preprocessing: Handled missing values, normalized numerical data, and encoded categorical variables.
• Model: Implemented a simple Linear Regression model.
• Evaluation: Measured performance using RMSE and R-squared scores.

2. XGBoost Regression (Iowa)

• Preprocessing: Feature engineering, handling missing values, and one-hot encoding categorical variables.
• Model: Used the XGBoost algorithm with hyperparameter tuning.
• Evaluation: Assessed model accuracy using RMSE and feature importance analysis.

Installation & Usage

1. Clone the repository:

   git clone https://github.com/turjoy18/housing-price-prediction.git
   cd housing-price-prediction

2. Install dependencies:

   pip install -r requirements.txt

3. Run the Jupyter Notebooks to explore the models:

   jupyter notebook

Results & Findings

• The Linear Regression model provided a baseline prediction for housing prices in Dumbo, Brooklyn.
• The XGBoost model showed improved accuracy, leveraging feature selection and boosting techniques for better predictive performance.

Future Improvements

• Incorporate additional features such as economic indicators and interest rates.
• Experiment with more advanced machine learning techniques like Neural Networks.

Author

turjoy18 - Passionate about Machine Learning and Data Science.