Dermatology Diagnosis Dataset with Nominal Attributes

BNG(dermatology,nominal,1000000) is a dermatology dataset from the OpenML platform, containing nominal attribute data. The dataset's name suggests a focus on skin disease classification using categorical features. Specific details on the creator, collection date, and exact row count are not provided in the available metadata.

Use Cases

• Classify skin disease types using nominal clinical and histopathological attributes.
• Train multi-class classifiers to predict patient age group or family history from symptom profiles.
• Benchmark feature selection algorithms on high-dimensional nominal medical data.
• Analyze correlations between nominal clinical features like erythema or scaling and specific diagnoses.

Strengths

• Dataset name indicates a nominal data structure suitable for categorical analysis.
• Sourced from OpenML, a known repository for machine learning datasets.

Limitations

• No confirmed row count, column names, or sample data available for validation.
• Lack of metadata on collection methodology, time range, and geographic origin limits reproducibility.
• Unknown license and author information restricts clarity on usage rights.

Provenance

Source

OpenML