diskriminatorer

Diskriminatorer (discriminators) are devices, circuits, or algorithms designed to distinguish among alternatives based on measured features. The term is used across engineering, statistics, and computer science to describe mechanisms that decide between categories or extract a signal corresponding to a class.

In electronics and signal processing, a discriminator is a circuit that converts a modulated signal into a usable representation. A frequency discriminator converts frequency variations into voltage to demodulate frequency-modulated signals; a threshold or limiter discriminator compares an input against a reference and produces a binary decision, as in many digital interfaces or safety circuits.

In statistics and machine learning, a discriminator is a model that assigns inputs to categories. In generative adversarial networks (GANs), the discriminator tries to distinguish real data from synthetic data. More generally, a discriminant function underpins methods such as linear discriminant analysis, support vector machines, and decision trees, which separate classes based on learned boundaries.

In social or everyday language, the term can refer to people or mechanisms that distinguish between groups or options, sometimes neutrally and sometimes with pejorative connotations when it leads to unfair treatment or bias.

See also: discriminant, detector, classifier.