Ames Housing Analytics
Ames, Iowa Housing Dataset 2006–2010 · 2,929 transactions 3 predictive models
Property Price Intelligence · Live Demo

What drives house prices
in Ames, Iowa?

Three models — from transparent formula to machine learning — tested on 586 unseen properties.

2,929
Properties analysed
28
Neighbourhoods
90.9%
Best model accuracy

Key Market Insights

From the data
Quality Premium · Current Model
$20,578
Each quality grade step adds $20,578 to the estimated price. The strongest single pricing driver across all models.
↑ Primary pricing driver
Location Premium · Current Selection
$56,261
NridgHt–MeadowV spread: $56,261. NAmes is $8,076 above MeadowV (lowest).
↑ 461% above lowest
Model Accuracy · Holdout Validation
90.9%
Random Forest explains 90.9% of price variation on 586 unseen properties. Typical error ±$26,253 per property.
↑ Strong predictive fit

Configure Property Attributes

Interactive
Quick scenarios:
Quality (1–10)
6
Floor Area (sqft)
1,444
Basement (sqft)
990
Garage Capacity
2
Year Built
1,973
Neighbourhood
Validated on 586 unseen properties · Results are illustrative estimates only

Model Estimates

Not yet run
Model 1 — Formula Transparent
Linear regression using 5 interpretable factors.

Configure attributes above and run models to see the estimate.
Model 2 — Formula + Location Location-Adjusted
Formula model extended with neighbourhood premium/discount.

Improves accuracy from 77% to 81%.
Model 3 — Random Forest Machine Learning
Ensemble of 100 decision trees — 32 features including neighbourhood.

Best accuracy: 90.9% on unseen data.

Session History

0 predictions
Session History
0 predictions
Model 1 — Formula
Model 2 — Formula+Location
Model 3 — ML
No predictions yet. Configure a property and run the models.