Metric | Value |
---|---|
R-squared | 80.8% |
Adjusted R-squared | 80.7% |
Mean Absolute Error | $21,655.09 |
Root Mean Squared Error | $34,819.79 |
To identify the key drivers of home values, we built a multiple linear regression model using the most influential features.
Metric | Value |
---|---|
R-squared | 80.8% |
Adjusted R-squared | 80.7% |
Mean Absolute Error | $21,655.09 |
Root Mean Squared Error | $34,819.79 |
Our model explains 80.8% of the variation in housing prices using these key features.
The model coefficients tell us exactly how much each feature impacts the home price, all else being equal.
Our model does a good job of predicting home prices, but let’s examine how it performs across different price ranges.
The true value of our model is in understanding the relative importance of features.
Here’s what standardized coefficients tell us about what really drives home prices:
A home’s overall quality rating is the single most consistent predictor of its price.
Despite quality and size being major factors, location continues to be a fundamental driver of home values.
Newer homes tend to command a premium. Homes built in the 2000s typically sell for $20,000 more than similar homes from the 1920s.
If you’re looking to increase your home’s value, kitchen upgrades and overall quality improvements give you the best bang for your buck.