About UCI Machine Learning Repository
The UCI Machine Learning Repository is a long-running library of datasets for machine learning. Students and researchers use it to test algorithms and compare results on data that many others have already used.
What it does
The site hosts datasets that have been donated or maintained for the machine learning community. Each entry describes the data, the number of instances and features, and the kind of task it suits, such as classification or regression. Users can open a dataset, read its notes and download it for their own work.
Key features
- Datasets tagged by task, such as classification
- Instance and feature counts for each entry
- Classic examples such as Iris and Wine Quality
- A way to contribute a new dataset to the community
Who it's for
Students learning machine learning, instructors building exercises and researchers who want a well-known benchmark dataset. It is useful for checking a model against data that other people have also tried.
Pricing and good to know
Datasets are free to browse and download. Each dataset comes with its own description and license terms, so check those before reuse. The site is moving to a new version, so some page addresses may change over time.
Highlights
- Find datasets tagged by task, such as classification
- View instance and feature counts for each dataset
- Download classic datasets such as Iris and Wine Quality
- Contribute a new dataset to the community
