Skip to content

README.md

Latest commit

 

History

History
139 lines (94 loc) · 3.94 KB

README.md

File metadata and controls

139 lines (94 loc) · 3.94 KB

🏡 Airbnb Price Predictor

Accurately predicting the nightly price of Airbnb listings using regression models and data preprocessing techniques.

📌 Table of Contents

🧠 Overview

This project explores price prediction for Airbnb listings through a regression modeling pipeline. Three models are evaluated:

  • Linear Regression
  • Gradient Boosted Regressor
  • Optimized Gradient Boosted Regressor (via GridSearchCV)

The end goal is to develop a model that can power an intelligent Price Estimation Tool for new Airbnb hosts.

⚙️ Pipeline Architecture

1. Data Preparation

  • Feature engineering: added and removed relevant features
  • Imputed missing values in 'beds' using pandas.isnull() and fillna()
  • Winsorized outliers in 'price' and 'review_scores_rating'
  • Applied log transformation to 'price' to normalize the distribution
  • Converted has_availability from boolean to numeric
  • One-hot encoded room_type and neighbourhood_group_cleansed
  • Computed a correlation matrix to identify key features
  • Dropped redundant columns like beds_na and original winsorized columns

2. Linear Regression

  • Split data into training and test sets
  • Applied StandardScaler for normalization
  • Fitted a LinearRegression model
  • Evaluated with RMSE and R²

3. Gradient Boosted Trees (GBDT)

  • Fitted a GradientBoostingRegressor with:
    • max_depth=10
    • n_estimators=300
  • Evaluated with RMSE and R²

4. Hyperparameter Tuning

  • Used GridSearchCV to optimize:
    • max_depth
    • n_estimators
  • Trained GBDT with optimal parameters
  • Compared performance against baseline models

📊 Results

Model R² Score RMSE
Linear Regression 0.42 104.9
GBDT (Non-Optimized) 0.56 91.4
GBDT (Optimized via GridSearchCV) 0.76 89.1

🚀 Future Work

  • Expand to larger, more diverse property datasets
  • Perform deeper EDA and feature scaling
  • Explore advanced models:
    • Neural Networks
    • Stochastic Gradient Descent Regressor (SGD)
  • Target a reduced RMSE of < 50

🎯 Project Purpose

Affordability is a core value of Airbnb. This model contributes to a smart Price Estimation Tool to assist new hosts in setting competitive and fair prices.

"Price too low? Lose revenue. Too high? Lose customers."

This model ensures:

  • Hosts get data-driven price estimates
  • Travelers benefit from fair pricing
  • Airbnb improves marketplace efficiency

🛠️ Tech Stack

📊 Machine Learning Models

  • Linear Regression — baseline model for price prediction
  • Gradient Boosted Decision Trees (GBDT) — for improved non-linear performance
  • GridSearchCV — for hyperparameter tuning and model optimization

🧹 Data Processing & Feature Engineering

  • Pandas — data manipulation and preprocessing
  • NumPy — numerical computing
  • Scikit-learn — preprocessing tools (e.g., StandardScaler, OneHotEncoder)

📈 Data Visualization

  • Matplotlib — basic plotting
  • Seaborn — statistical visualizations (e.g., correlation heatmaps)

🧪 Experimentation & Development

  • Jupyter Notebook — interactive experimentation environment
  • Python 3.8+ — core programming language

📫 Let’s Connect!
If you found this project insightful or want to collaborate, feel free to reach out or check out my GitHub profile.