We publish our scoring methodology because investors deserve to understand how recommendations are made.
Median sale price, days on market, homes sold, list-to-sale ratio, and inventory levels across 200+ zip codes updated monthly.
Median household income and bachelor's degree attainment rates by zip code from public demographic data sources.
Fair market rent for 2-bedroom units by metro area from HUD published data, updated annually.
24 months of historical median sale price data per zip code enabling trend analysis, velocity calculation, and year-over-year comparisons.
| Feature Name | What It Measures |
|---|---|
| YoY Appreciation | Annual price change % vs 12 months ago |
| Price Velocity | 3-month rolling average of monthly price change |
| Market Heat Index | Composite of DOM, volume, ratio, appreciation |
| Gross Rental Yield | Annual rent / sale price × 100 |
| Net Rental Yield | Gross yield after 25% expense assumption |
| Neighborhood Score | Weighted income (40%) + education (30%) + home value (30%) composite |
These weights are fixed and disclosed. We do not change them without announcement.
All six components normalized 0-100 using min-max scaling before weighting. Final score ranges from 0 to 100.
A Random Forest regression model trained on time-ordered historical data (80% train, 20% test, no shuffle) to predict median sale price 6 months forward. The model uses five input features: appreciation, velocity, market heat, neighborhood score, and rental yield.
A weighted composite model scoring each zip code 0-100 across six investment factors with publicly disclosed weights. Not a supervised ML model — a principled scoring framework updated monthly.
K-Means clustering (k=5) groups zip codes into five market personality types based on appreciation, heat index, rental yield, neighborhood score, days on market, and price per sqft. Cluster labels are derived programmatically from centroid values.
IMPORTANT DISCLOSURE: All market analysis generated by the Albert Realty Group Texas AI Investment Intelligence Platform is for informational purposes only. Market data used in this platform is synthetic data generated for demonstration purposes. This is not financial advice. Past performance does not guarantee future results. Always conduct your own due diligence before making any investment decision.