exemplarbaserte

Exemplarbasert is a Norwegian term for approaches or theories that are based on exemplars, i.e., concrete examples, rather than abstract generalization. In cognitive science, exemplar-based theories argue that memory and recognition rely on comparisons with stored specific instances rather than a single abstract prototype. In machine learning and information retrieval, exemplar-based methods store a set of observed cases (exemplars) and predict or reason by comparing new inputs to these exemplars, often using similarity measures.

Exemplar-based learning is often contrasted with prototype-based approaches, where a prototype summarizes a class. Common exemplar-based techniques include k-nearest neighbors in supervised learning and case-based reasoning in AI, which reason from past cases to solve new problems.

Advantages include simplicity, interpretability, and flexibility with irregular data; limitations include storage requirements, sensitivity to noisy exemplars, and slower prediction at scale compared to model-based methods.

Within Norwegian usage, exemplarbasert is used across disciplines such as psykologi, maskinlæring, and informasjonsgjenfinning, sometimes with anglicisms like exemplarbasert læring or exemplarbasert hukommelsesteori. See also: case-based reasoning, k-nearest neighbors, prototype theory, exemplar theory.