In Artificial intelligence, the method of the K closer neighbors is a method of Apprentissage supervised.

Within this framework, one lays out of a Database of training consisted of NR couples “input-output”. To consider the exit associated with a new entry X , the method of the K closer neighbors consists in taking into account (in an identical way) the K samples of training to which the entry is closest to the new entry X , according to a distance to be defined.

For example in a problem of classification, one will retain the class most represented among the K left associated with the K entered closest to the new entry X .

See too

• Research of the closest neighbors