Les Perceptrons Multicouches: de la régressionnon-linéaire aux problèmes inverses
Badran, Fouad ; Thiria, Sylvie
HAL, Report N°: CEDRIC-01-242 / Harvested from HAL
Text on HAL
The problem of nonlinear least square regression using multi-layeredperceptrons is addressed inthispaper. A general formulation usingmaximum likelihood is given and a particular attention is paid toalgorithm accounting for uncertainties. The generalization to densityestimation is given. In the second part of the paper, a generalframework allows to modelize general inverse problems.
Publié le : 2001-07-04
Classification:  [INFO]Computer Science [cs],  [MATH.MATH-ST]Mathematics [math]/Statistics [math.ST] 
@article{Report N°: CEDRIC-01-242,
     author = {Badran, Fouad and Thiria, Sylvie},
     title = {Les Perceptrons Multicouches: de la r\'egressionnon-lin\'eaire aux probl\`emes inverses},
     journal = {HAL},
     volume = {2001},
     number = {0},
     year = {2001},
     language = {en},
     url = {http://dml.mathdoc.fr/item/Report N°: CEDRIC-01-242}
}

Badran, Fouad; Thiria, Sylvie. Les Perceptrons Multicouches: de la régressionnon-linéaire aux problèmes inverses. HAL, Tome 2001 (2001) no. 0, . http://gdmltest.u-ga.fr/item/Report%20N%C2%B0:%20CEDRIC-01-242/