Dark Mode
Utrecht Housing Price Prediction Data
Comodities & Real Estate
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
The prediction of residential property prices based on key factors such as location, size, and other physical attributes. It provides an excellent tool for those learning data science and machine learning principles, particularly focusing on responsible AI by making algorithm testing and evaluation accessible to a wide audience. Unlike many existing academic datasets, this material offers meaningful tasks and avoids common data quality issues that can distract from educational objectives. The information compiled here is derived from actual housing offer details (Funda) and official Dutch land registry records (Kadaster).
Columns
- id: A unique numerical identifier for each property, ranging between 5000 and 9000.
- zipcode4: The four digits of the postal code, generally representing a sub-district area.
- zipcode6: The complete six-character (four digits, two letters) Dutch postal code, corresponding to a specific part of a street.
- zipcode6id: A unique identifier created by appending three digits to the
zipcode6. - housetype: Indicates the dwelling type, either
woonhuis(residential house) orappartement(flat). - lot-area: The total area of the land plot in square metres. Flats (
appartementtypes) are typically recorded with 0 lot-area. - house-area: The living area of the house in square metres.
- garden-size: The size of the garden in square metres.
- rooms: The total number of rooms in the property.
- bathrooms: The number of bathrooms, with most properties featuring one or two.
- x-coor: Latitude (decimal degrees) of the house position, rounded to four decimals for precision.
- y-coor: Longitude (decimal degrees) of the house position, rounded to four decimals.
- buildyear: The year of construction. While most homes were built in the 20th and 21st century, the oldest house recorded dates back to 1320.
- retailvalue: The final transaction value in thousands of euros, officially registered in the Dutch land registry.
- askingprice: The price at which the house was offered in 2024.
- energylabel: The EU energy efficiency rating, ranging from G (not energy efficient) to A++ (highly efficient).
- energyeff: A binary variable (1 or 0) indicating if the house is considered very energy efficient (Energy label A or higher).
- valuationdate: The official transaction date when the final value was registered, following the ISO 8601 format.
- street: The name of the street.
- subdistrict: One of 33 sub-districts or neighbouring towns within the Utrecht region.
- district: One of 10 major Utrecht districts or neighbouring towns.
- city: The local township (
gemeente), primarily Utrecht, but including close sub-urban areas like Vleuten. - dist-from-train: The 'crow's flight' distance in kilometres from Utrecht central train station.
Distribution
The data is provided as a CSV file (
2025-housing-dataset-alldata.csv), where each line represents one property and values are separated by commas. The file size is 23.65 kB. The dataset contains 153 records, with virtually all columns being fully populated.Usage
The primary use case is developing and testing algorithms for real estate valuation and price prediction. It is particularly well-suited for classroom instruction and visualisations for students learning AI and machine learning, allowing them to train and test algorithms with only basic training.
Coverage
The geographical scope focuses on Utrecht and its immediate neighbouring towns in the Netherlands. The time range for the official valuation dates spans from January 2024 through to November 2024, with asking prices specifically reflecting 2024. The data provides historical context through property construction years, which range from the 14th century up to 2019.
License
CC BY-SA 4.0
Who Can Use It
This material is ideal for students and academic researchers focusing on machine learning, data science, or urban economics. It enables educators to illustrate complex phenomena using a practical, meaningful dataset. Data enthusiasts interested in Dutch real estate market metrics will also find it valuable.
Dataset Name Suggestions
- Utrecht Housing Price Prediction Data
- Dutch Housing Appraisal Dataset 2024
- Utrecht Real Estate Market Indicators
- Netherlands Housing Valuation Data
Attributes
Original Data Source:Utrecht Housing Price Prediction Data
QUALITYLoading...