# App of GenAI/Deep Learning (WUSTL, Fall 2025) — Data

来源：https://www.kaggle.com/competitions/app-of-gen-ai-deep-learning-wustl-fall-2025/data

采集时间：2026-10-19T07:40:12Z

## Competition Description
This competition challenges participants to predict property value scores for residential properties. The dataset contains detailed information about houses, including both exterior and interior characteristics, along with an accompanying exterior image for each property.

## Data Description
The dataset provides comprehensive information about residential properties, including architectural details, materials, dimensions, and various features. Each property is represented by both structured data and an exterior photograph (named as `[id].jpg`).

## Evaluation
Submissions are evaluated using Root Mean Square Error (RMSE) between predictions and actual property value scores.

### Files
- `train.csv` - Training set
- `test.csv` - Test set
- `sample_submission.csv` - Sample submission file in the correct format
- Image files corresponding to each property (e.g., "1.jpg" for property with id=1)

### Data Dictionary

| Column Name | Type | Description |
|-------------|------|-------------|
| id | int | Unique identifier for each property |
| architectural_style | string | Style of the house (e.g., modern, colonial, craftsman) |
| exterior_material_primary | string | Main material used on exterior (e.g., brick, wood, stucco) |
| environment_type | string | Setting of the property (coastal, urban, suburban, rural, mountain) |
| construction_year | int | Year the property was built |
| total_area_sqm | float | Total area in square meters |
| bedroom_count | int | Number of bedrooms |
| bathroom_count | float | Number of bathrooms |
| garage_capacity | int | Number of cars the garage can hold |
| energy_rating | string | Energy efficiency rating (A-G) |
| smart_home_level | string | Level of smart home integration |
| landscape_style | string | Style of landscaping |
| interior_style | string | Interior design aesthetic |
| floor_material_primary | string | Main flooring material |
| wall_material_primary | string | Main wall material |
| lighting_type | string | Primary lighting style |
| story_count | int | Number of stories |
| exterior_color_primary | string | Main exterior color |
| interior_color_palette | string | Interior color scheme |
| home_type | string | Type of residence (detached, townhouse, etc.) |
| property_value_score | float | (Target) Property value score (0-100) |

Note: The target variable (property_value_score) is only provided for the training set.

## Image Data
Each property includes a high-resolution exterior photograph. Images are provided in JPEG format and named according to their corresponding property ID (e.g., "1.jpg" for the property with id=1).