TERMIUM Plus^®

Fiche 1,  Justifications,  Anglais

Record number: 1, Textual support number: 1 DEF

A projection technique that reduces data dimensionality, and an advanced neural network that utilizes fuzzy-set mathematics.  3, fiche 1, Anglais, - n%2Ddimensional%20probability%20density%20function

Record number: 1, Textual support number: 1 CONT

The intent for multispectral and hyperspectral image data analysis is to rapidly and inexpensively associate a ground cover label to each pixel in the image. Given the multivariate nature of such data, the process of data analysis is one of dividing up the N-dimensional feature space into M exhaustive but non-overlapping regions where M is the number of classes of materials existing in the scene. The process involves defining the M classes of interest in a quantitative fashion, such that each pixel in the scene, which exists as a discrete location in the N-dimensional space, can be uniquely associated with one of the M classes. Frequently, this is done by using a small number of samples in the scene, called design samples or training samples, to define an N-dimensional probability density function for each of the M classes.  4, fiche 1, Anglais, - n%2Ddimensional%20probability%20density%20function

Fiche 1,  Justifications,  Français

Record number: 1, Textual support number: 1 DEF

Technique de projection qui diminue la dimensionnalité des données, et réseau neuronal perfectionné qui utilise les mathématiques floues.  1, fiche 1, Français, - densit%C3%A9%20de%20probabilit%C3%A9%20en%20n%20dimensions

Fiche 1,  Justifications,  Espagnol