Airbnb Project
Published:
The Main Work
- Applied Python to manipulate EDA and feature engineering, solve missing value by KNN and K-means, extract newly important holidays, weekend, travel months, location(zip codes), and sentiment features
- Predicted prices with machine learning methods including(Xgboost, Random Forest Regression, Lasso, Ridge), achieved 87% accuracy with Xgboost
- Proposed recommendations for Airbnb hosts on being a super host, holding a flexible cancelling policy, providing detailed house pictures, verification and instant bookable rooms tend to attract more customers