Housing Price Prediction Project

Overview

This repository contains the data processing, predictive modeling workflows, and the implementation of a residential housing price estimation tool.

Project Structure

1. Data Merging

• Workflow: Combines numerical and categorical engineered datasets into a unified master file.
• Key Files: numerical_engineered_data.csv, categorical_engineered_data.csv
• Output: final_model_ready_data.csv

2. Price Prediction Calculator

• Tool: A custom-built estimation tool featuring a predict_house_price function.
• Functionality: Takes user-defined property features (e.g., quality, square footage, year built) as input.

3. Random Forest Modeling

• Notebook: Focuses on establishing a baseline using a RandomForestRegressor.
• Performance: Achieved an $R^2$ of approximately 0.89.

4. Ridge Regression Modeling

• Notebook: Focuses on linear modeling techniques.
• Preprocessing: Implementation of a Pipeline including SimpleImputer and StandardScaler.
• Optimization: Applied GridSearchCV for alpha tuning.
• Performance: The tuned Ridge model with log-transformed targets achieved an $R^2$ of ~0.91.

Tableau Presentation

The findings and the predictive logic are being visualized in a Tableau dashboard to communicate regional insights and model performance to stakeholders.# house-prices-project