The objective of this project is to build price suggestion models which might help hosts in getting an idea of the fair price of their house. Exploration and visualization aids in feature extraction for developing predictive models using different statistical learning algorithm. Our Predictive models aim to forecast the listing price on a given day of the year.
Predictive and visualization models like these are interesting for many stakeholders: -
- Airbnb Hosts: Hosts can identify competitive price for their house and give them a chance to improve their hospitality
- Customers: Anybody planning to book a hotel wants to get a competitive price with the maximum number of features. My price suggestion model will provide them a fair price on the listing of their choice.
- Data scientists: As budding data scientists we always try to learn from other models and think about ways on how to improve the existing model, this helps in eliminating the need of reinventing the wheel.
The primary data source comes from the publicly available Airbnb data website. I used listings, reviews and calendar data available on the website.
In addition, traffic data from ---- was incorporated into the Cab data for prediction modeling.
Python
- Python Standard Library: Built in python modules.
- Numpy: Scientific computing with python.
- Matplotlib: Python plotting.
- Seaborn: Data visualization built on Matplotlib.
- Pandas: Data analysis tools.
- Scikit-learn: Machine learning toolkit.
R
- dplyr: A Grammar of Data Manipulation
- data.table: Fast aggregation of large data
- anytime: Anything to 'POSIXct' or 'Date' Converter
- lubridate: Functions to work with date-times and time-spans