- perform eda on data for a better understanding
- build deep learning model to predict the price of houses
- Construct API
- Host API using flask
The objective was based on the kaggle data set - https://www.kaggle.com/brittabettendorf/berlin-airbnb-data.
- Deep Learning
- Machine Learning
- CNN neural network modeling
- EDA
- Python
- Pandas, GoogleColab
- Numpy
- Matplotlib
- Torch
When listing houses on airBnB the question investors ask the most is often how do you maximise profit given the supply of competition as well as the demand by customers. This project seeks to bring a more systematic method of determining prices of houses for investors.
listings are as of November 2018.
- Most rooms offered are private rooms or entire apartments
- Most beds offered are real beds as opposed to futons or couches
- Most reviews are 8/9/10s with a few exceptions
- Most places are instantly bookable but not by an overwhelming majority
- It is almost a must for hosts to have profile pictures
- The highest property was valued at $999 (does not specify currency)
- Businesses who are in the business of real estate who want to capitalise on maximising ROA can use this tool to aid them their the decision making.
- Persons who own property in Berlin have a tool to estimated how much their property is worth if they wanted to rent it out with airbnb
Team Leads (Contacts) : [Samuel Lawrence]: http://samuel-lawrence.co.uk/