median_house_value_usdOpenML #537 ↗Houses
The California Houses dataset contains 20,640 census block group records from the 1990 U.S. Census, structured to predict median house values using demographic, spatial, and structural real estate features.
Business Objective: Houses
01Business Context
Automated valuation models (AVMs) and property investment firms require granular, location-sensitive appraisals to assess neighborhood residential asset values accurately without manual site inspections.
02Analytical Objective
Predict the median residential house value for geographic block groups based on spatial coordinates, household demographics, and structural housing metrics.
03Economic & Decision Impact
Accurate property valuation mitigates underwriting risk in residential mortgage lending, eliminates appraisal latency, and prevents severe capital misallocation stemming from under- or over-valued real estate portfolios.
Tree-based ensembles such as LightGBM, XGBoost, and CatBoost historically establish strong baselines on the California Houses dataset by capturing non-linear geographical interactions between latitude and longitude. Zero-shot tabular foundation models like TabICL and Carla match or exceed fine-tuned gradient-boosted trees in few-shot in-context settings without requiring hyperparameter tuning or feature engineering.
Target & Feature Column Definitions
Exhaustive business definitions, measurement units, roles, and target variables across all 9 columns.
| Col | Variable / Feature Name | Role | Type | Unit / Scale | Description & Business Meaning |
|---|---|---|---|---|---|
| A | median_house_value_usdMedian House Value | TARGET | NUM | USD | Median house value for households within the census block group |
| B | median_income_tens_of_thousandsMedian Income | Feature | NUM | 10k USD | Median household income within the census block group (expressed in tens of thousands of USD) |
| C | housing_median_age_yearsMedian Housing Age | Feature | NUM | years | Median age of residential buildings within the census block group |
| D | total_rooms_countTotal Rooms | Feature | NUM | count | Total count of rooms across all residential units in the block group |
| E | total_bedrooms_countTotal Bedrooms | Feature | NUM | count | Total count of bedrooms across all residential units in the block group |
| F | population_countPopulation | Feature | NUM | people | Total number of permanent residents residing within the census block group |
| G | households_countHouseholds | Feature | NUM | households | Total number of households (occupying housing units) in the block group |
| H | latitude_degLatitude | Feature | NUM | degrees | Centroid latitude coordinate of the census block group |
| I | longitude_degLongitude | Feature | NUM | degrees | Centroid longitude coordinate of the census block group |
Interactive Data Table
Explore rows, feature values, and target labels for Houses.
TabICLv2 WebGPU Benchmark Results
Complete performance metrics on persisted IID or non-IID evaluation splits, executed 100% locally in the browser sandbox.
| Evaluation Split | Train Rows | Test Rows | Split Duration?Wall-clock inference time taken for this evaluation split running 100% locally via WebGPU.Learn more → | R² Score?Coefficient of determination measuring variance explained by the model.Learn more → | MAE?Mean Absolute Error in original target units, robust to outliers.Learn more → | RMSE?Root Mean Squared Error, penalizing large residual deviations.Learn more → | MAPE?Mean Absolute Percentage Error relative to true target values.Learn more → | RMSLE?Root Mean Squared Logarithmic Error, measuring proportional error scales.Learn more → |
|---|---|---|---|---|---|---|---|---|
| R1 / F1 | 13,760 | 6,880 | 50.33s | 0.8727 | 25,805.59 | 41,603.12 | — | — |
| R1 / F2 | 13,760 | 6,880 | 49.97s | 0.8604 | 26,207.94 | 42,515.54 | — | — |
| R1 / F3 | 13,760 | 6,880 | 50.03s | 0.8633 | 26,266.76 | 42,789.29 | — | — |
| R2 / F1 | 13,760 | 6,880 | 49.99s | 0.8704 | 25,515.93 | 41,399.63 | — | — |
| R2 / F2 | 13,760 | 6,880 | 50.31s | 0.8646 | 26,425.72 | 42,777.59 | — | — |
| R2 / F3 | 13,760 | 6,880 | 50.45s | 0.8694 | 25,463.26 | 41,516.92 | — | — |
| R3 / F1 | 13,760 | 6,880 | 50.46s | 0.8720 | 25,459.16 | 40,838.46 | — | — |
| R3 / F2 | 13,760 | 6,880 | 50.41s | 0.8573 | 26,925.39 | 43,289.24 | — | — |
| R3 / F3 | 13,760 | 6,880 | 50.31s | 0.8647 | 25,859.79 | 43,191.18 | — | — |
| Mean ± Std | — | — | 50.25s | 0.8661 ± 0.0050 | 25,992.2 | 42,213.4 | — | — |
Foundation Models vs Traditional ML Baselines
Side-by-side evaluation of our in-browser Carla engine (WebGPU), open-source tabular foundation models, and OpenML baselines.
| Rank | Algorithm / Model | Model Family | Runtime | R² Score?Coefficient of determination (R²) measuring proportion of variance explained.Google ML Guide & Details → | MAE?Mean Absolute Error measuring average prediction error magnitude.Learn more → | RMSE?Root Mean Squared Error penalizing large prediction errors.Learn more → | Reference / Repo |
|---|---|---|---|---|---|---|---|
| #1 | Google TabFM v1.0 | Tabular Foundation Model | PyTorch/Python | 0.8944 | 22,351.1 | 37,481 | google-research/tabfm ↗ |
| #2 | TabPFN v3 | Tabular Foundation Model | PyTorch/Python | 0.8826 | 23,749.7 | 39,522.7 | PriorLabs/tabpfn ↗ |
| #3 | EXAONE Tabular | Tabular Foundation Model | PyTorch/Python | 0.8800 | 24,044.7 | 39,970.9 | LGAI-Research/EXAONE-Tabular ↗ |
| #4 | Full TabICLv2 (PyTorch Reference) | Tabular Foundation Model | PyTorch/Python | 0.8739 | 24,947.3 | 40,970.9 | soda-inria/tabicl ↗ |
| #5 | nanotabicl Vanilla | Tabular Foundation Model | PyTorch/Python | 0.8737 | 24,965.6 | 41,005.3 | soda-inria/nanotabicl ↗ |
| #6 | Streaming nanotabicl | Tabular Foundation Model | PyTorch/Python | 0.8717 | 25,242.7 | 41,314.9 | soda-inria/nanotabicl ↗ |
| #7 | Carla Engine (TabICLv2 WebGPU) | Tabular Foundation Model | Browser | 0.8661 | 25,992.2 | 42,213.4 | 100% In-Browser |
Variable Schema & Summary Distributions
Observed numerical ranges, category cardinalities, missing rates, and sample values across 20,640 rows.
| Col | Variable Name | Type | Missing | Distinct | Summary Stats / Distribution | Sample Values |
|---|---|---|---|---|---|---|
| A | median_house_value_usdTARGET | NUM | 0 | 3,842 | [14999, 500001] μ=206855.8 σ=115392.8 | 452600358500352100 |
| B | median_income_tens_of_thousands | NUM | 0 | 12,928 | [0.4999, 15.0001] μ=3.9 σ=1.9 | 8.32528.30147.2574 |
| C | housing_median_age_years | NUM | 0 | 52 | [1, 52] μ=28.6 σ=12.6 | 412152 |
| D | total_rooms_count | NUM | 0 | 5,926 | [2, 39320] μ=2635.8 σ=2181.6 | 88070991467 |
| E | total_bedrooms_count | NUM | 0 | 1,928 | [1, 6445] μ=537.9 σ=421.2 | 1291106190 |
| F | population_count | NUM | 0 | 3,888 | [3, 35682] μ=1425.5 σ=1132.4 | 3222401496 |
| G | households_count | NUM | 0 | 1,815 | [1, 6082] μ=499.5 σ=382.3 | 1261138177 |
| H | latitude_deg | NUM | 0 | 862 | [32.54, 41.95] μ=35.6 σ=2.1 | 37.8837.8637.85 |
| I | longitude_deg | NUM | 0 | 844 | [-124.35, -114.31] μ=-119.6 σ=2.0 | -122.23-122.22-122.24 |
Frequently Asked Questions: Houses
Common questions regarding the Houses dataset, machine learning task formulations, and in-browser tabular inference.
What is the Houses dataset used for?
The Houses dataset is a classic tabular regression benchmark used to predict median residential home values in California based on 1990 census spatial and demographic indicators.
Can I run zero-server predictions on this dataset in Google Sheets?
Yes, Carla HQ enables zero-server in-context tabular prediction directly in spreadsheets using TabICL foundation models.
What machine learning models perform best on Houses?
Gradient-boosted decision trees (XGBoost, LightGBM, CatBoost) and tabular foundation models like TabICL perform best, capturing complex non-linear spatial relationships across latitude and longitude coordinates.
Test TabICLv2 on Houses Yourself
Open the pre-loaded Google Sheet and let Carla configure the target and task for local, zero-cloud tabular machine learning.
Launch Carla
Click once to open the spreadsheet and Carla side panel together.
Review the Setup
Carla selects the dataset target and task from this page automatically.
Evaluate & Predict
Run predictions and compute metrics with zero server uploads.
Provenance & Attribution
OpenML Dataset 537