Authors: Lorela Blaka, Meaghan Ross, Rashid Karriti
This project analyzes a series of linear regressions to provide recommendations for King County Real Estate to have a better understanding of how to price homes in Washington's King County. Interpretive analysis shows that income per capita and the square footage of the home are strong indicators of how to price a home.
King County Real Estate has contracted us to investigate what home features have the greatest impact on pricing and speeding up the sale of the home. Based on our models, King County Real Estate should:
A) Review the relationship between the square footage of the home and price of a home and
B) Review the relationship between income per capita of a neighborhood and price of a home.
The project utilizes a descriptive analysis, linear modeling, and running several iterations of an OLS regression on home features to reveal which factors contribute most to housing prices. Our main data sets are King County House Sales Data from 2014 to 2015, and US Census Bureau data on income per capita across cities in King County. The prediction accuracy of our varaibles included in our model accounted for about 67% of the variation in price, with a price prediction of +/- $201,491 from the actual price.
Here the Price vs Square Feet of the home presents a strong correlation between the size of the home and an increase in the price. Our recommendation is to price homes based on the number of the square footage of the home.
Here based on Mean Price by income shows a strong correlation between the income per capita of a neighborhood and an increase in price. Our recommendation is to price homes that have a higher income per capita.
Real estate companies working on pricing of homes in King County should consider the square footage of the home, and income per capita of the home to give the most accurate pricing evaluation.
For a more comprehensive evaluation of pricing a home in the future, a real estate company should further inquire into:
- Expanding the data to years beyond 2015, to see if pricing has changed over time with other important features to the home.
- Include property tax information to better predict prices in the future.
- Include more features of a neighborhood, such as a school district & public transit.
See the full analysis in the Jupyter Notebook or review this presentation.
Lorela Blaka: lblaka@gmail.com
Meaghan Ross: mer423@nyu.edu
Rashid Karriti: rhk48@georgetown.edu
├── README.md
├── data <-- CSV
├── Individuals Notebooks <--- Directory for individual workspaces
│ ├── meaghan
│ ├── lorela
│ ├── rashid
│
├── RealEstateProject.pdf <-- non-technical presentation slides
├── final_notebook.ipynb <-- Jupyter Notebook containing codes detailing project's analysis
└── .gitignore