duygut / hotel_booking_cancelation_with_tree_based_algorithms

Comparing different tree-based algorithms to find the best model for cancelation prediction

Kaggle - Hotel Booking Demand Data

Every year, more than 140 million bookings made on the internet and many hotel bookings made through top-visited travel websites like Booking.com, Expedia.com, Hotels.com, etc. According to Google data, hotels are booked in advance of 12 weeks.

This tutorial is in an IPython Notebook for Kaggle Dataset, hhttps://www.kaggle.com/jessemostipak/hotel-booking-demand

You can find the scripts for choosing the best predictive model for cancelation prediction over Hotel Booking Demand dataset.

There are 6 topics in the notebook.

1. Data Information
2. Exploratory Data Analysis and Feature Engineering
3. Dealing with Missing Data and Correlation Matrix
4. Hyperparameter Tunning and Feature Importance
5. Model Building
6. Classification Reports and Classification Matrix

Dependencies: Pandas NumPy SciKit-Learn Seaborn Matplotlib XGBoost