Parameter

A parameter is a quantity that defines a system, model, or function. It is fixed for a particular instance of analysis, but may vary across different scenarios. In many contexts parameters determine the shape, scale, or behavior of a relation, while the elements that are allowed to change in response to input are called variables or arguments.

In mathematics, parameters index a family of objects. A parameterization expresses a family as a function of one or more parameters. Parametric equations describe a curve by parameters; the set of possible parameter values is the parameter space. In statistics, a population parameter is an unknown characteristic of a population, such as the mean or variance; estimators compute approximate values from data, and properties like unbiasedness or consistency describe estimator quality.

In programming, a function or method has parameters in its signature; when invoked, actual values are passed as arguments. Parameter passing can be by value or by reference; some languages support default, named, or variadic parameters.

In physics and engineering, parameters include constants that define a model (for example, mass, resistance, or stiffness) and tuning knobs used in calibration. Parameter identification or calibration estimates these quantities from data, and sensitivity analysis studies how changes in parameters affect outcomes.

In machine learning, the term parameter often refers to learned quantities such as weights and biases; hyperparameters govern the learning process and are set prior to training.