Module 4 Final Project - Real Estate Price Forecasting

Purpose:

For this project, I will be answering the following question in order to consult a fictional real-estate investment firm:

What are the top 5 best zipcodes for us to invest in?

I will forecast real estate prices of various zipcodes using data from Zillow and recommend the top 5 best zipcodes to invest in.

Data Science Process Used:

I leveraged the OSEMN (Obtain, Scrub, Explore, Model, Interpret) process for this project. My notebook is organized to follow this process.

The Dataset:

The data I'm working with is a modified version of the data from the Zillow Research Page. The data file can be found in this repo as zillow_data.csv.

Key Question Answered:

What are the top 5 best zipcodes for us to invest in?

To qualify as one of the 'best' regions, the region must meet the following qualifications:

1. Have above average annual growth rate since the housing market recovered from the crisis (2012), and also have above average annual growth rate during the crisis (2007-2012)
2. Needs to have a 5yr average annual growth rate in the top 25% of the dataset
3. Needs to also have a 10 yr average annual growth rate in the top 25% of the dataset
4. Needs a narrow predicted interval width to ensure a more accurate forecasted value (interval width must be within 25% of the smallest interval widths
5. The maximum p-value must be less than alpha=.05 to ensure we are statistically significant and therefore a better performing region.