Data Objects and Attributes Overview

Study Notes

Data sets are comprised of data objects, which represent entities.
Examples of entities include customers, products, and sales in a sales database; patients and treatments in a medical database; and students, professors, and courses in a university database.
Data objects are also called objects, data points, examples, instances, records, samples, or tuples.
Attributes describe data objects.

An attribute is a data field that describes a characteristic or feature of a data object.
Examples of attributes are age, name, gender, and address.
The term dimension is often used in data warehousing, while machine learning literature uses feature and statisticians use variable.
The term attribute is commonly used by data mining and database professionals.
Observed values for a given attribute are known as observations.
An attribute/feature vector is a set of attributes that are used to describe a given object.

Nominal: Categories without ordering.
- Examples: color (black, blond, brown, grey, red, white) and marital status (single, married, divorced, widowed)
Binary: Attributes with two states, 0 and 1, such as gender (male, female).
- Symmetric binary: Both outcomes are equally important.
- Asymmetric binary: Outcomes have different importance, such as medical test results (positive, negative).
Ordinal: Values with order but not necessarily equal intervals, such as sizes (small, medium, large) and grades (A+, A, B+, B, C+, C, D, F).

Interval: Values are measured on a scale with equal units but lack a true zero point.
- Examples: temperature, calendar dates, intelligence scores, and blood pressure.
Ratio: Values have a true zero point, allowing for comparisons of magnitudes.
- Examples: length, counts, monetary quantities, and temperature in Kelvin.