/AirBnb

Exploring the Airbnb London dataset to find any interesting information as well as predicting the rent prices of different accommodations listed on the Airbnb platform in London

Primary LanguageJupyter NotebookMIT LicenseMIT

AirBnb: Project Overview

Brief introduction

  • Airbnb is a home-sharing platform and is popular in major cities such as London, New York, Amsterdam,Barcelona.
  • Hosts set a price when listing their houses on the platform
  • Due to heavy business competition in major cities, it is important for a business to set the right price for a product because a small difference in price could result in losing customers to your competitors.
  • AirBnb hosts have to endure this stiff competition hence the need to set the best price inorder to get more customers as well as providing the best customer service

Project Aims

The project will look at meeting the following goals:

  1. Which areas are mainly flocked by guests?
  2. Why are some hosts are busier than others?
  3. Why do guests highly rate some properties?
  4. Clean the dataset and prepare it for exploration and prediction
  5. Explore the dataset to find interesting relationships between.
  6. To predict the rent price of homes listed using machnine learning and deep learning
  7. To meet other project aims using a dataset from one of the following major cities: London, NewYork, Paris, San Francisco, Amsterdam, Barcelona, Berlin.

Code and Resources used

  • Python Version: 3.7
  • Jupyter notebook: 5.8.0
  • Packages: pandas,numpy, scikit-learn, matplotlib, seaborn, bokeh, holoviews(geoviews), geopandas, Keras, xgboost

Data Sources

The dataset for London was used and was obtained from Inside Airbnb

Exploratory Data Analysis findings

  1. Inner London boroughs were the most popular locations amongst guests with Westminster being the most popular borough in London.Borough popularity London1 Popularity
  2. Apartments are the commenest property types listed by hosts in London Apartments most common
  3. Guests preferred reserving entire houses/apartments compared to just a single room piechart
  4. The median price for homes was highest in the inner London boroughs and south weatern boroughs Median accommodation price
  5. The busiest hosts had more experience, provided more amenities and had properties in the most popular locations.
  6. 7 in 10 properties had a score of 80 and above which goes to show that most AirBnb users enjoy the service.
  7. Guests were more likely to leave a review on a property that offered them great service.

More findings can be found in the notebook files uploaded