The purpose of this project is to create a price prediction app for Airbnb NYC housings through various inputs such as borough, room type, reviews per month, and availability in a year.

You can access the application through the link: https://jdkwak1994-airbnb-price-predic.herokuapp.com

• https://www.kaggle.com/dgomonov/new-york-city-airbnb-open-data?select=AB_NYC_2019.csv
• https://www.kaggle.com/kritikseth/us-airbnb-open-data#__sid=js0
• http://insideairbnb.com/get-the-data.html

Step 1: Data Cleaning and Machine Learning Models

• Clean up the given data by adding 0 to any N/A data.
• Dummy data with neighborhood type, room type, reviews per month, and availability in 365.
• 3 models are created: linear regression model, xgboost model, and random forest model.

Step 2: HTML/CSS/JS

• Create application webpage under index.html
• Add different pages such as infographic, project information, and prediction.
• Make sure the pages are mobile friendly.

Step 3: Flask App and Heroku Deployment

• Connect the created html pages to app.py.
• Deploy the application through Heroku to make it accessible by public.

• Main Page

• Prediction Page

• Infographic Page

Steps below will run the application via Flask, which uses all 3 models

1. Clone this repo.
2. Uncomment (get rid of the #) lines 69 and 70 of app.py.
3. Comment out (add the # in front of the line) line 71.
4. Change the line 77 to avgprice = round((sum(price) / 3), 2).
5. Run flask run or python -m http.server or any other method for this purpose

• Jeongdae (JD) Kwak
• Adriana Cieslak
• Anumala Thapa
• Samuel Okunola