DD_tools: autoenc_dd

AUTOENC_DD Auto-Encoder data description.
 
       W = AUTOENC_DD(A,FRACREJ,N)
 
 Train an Auto-Encoder network with N hidden units. The network should
 recover the original data A at its output. The difference between the
 network output and the original pattern (in MSE sense) is used as a
 charaterization of the class. The threshold on this measure is optimized
 such that FRACREJ of the training objects are rejected.
 
 Default: N=5

See also: |datasets| |mappings| dd_roc 

@phdthesis{Tax2001a,
	author = {Tax, D.M.J.},
	title = {One-class classification},
	school = {Delft University of Technology},
	year = {2001},
	address = {http://www.ph.tn.tudelft.nl/\~{}davidt/thesis.pdf},
	month = {June}
}



Copyright: D.M.J. Tax, D.M.J.Tax@prtools.org
Faculty EWI, Delft University of Technology
P.O. Box 5031, 2600 GA Delft, The Netherlands