TY - CHAP
T1 - Conclusions and future research
AU - Kolman, Eyal
AU - Margaliot, Michael
PY - 2009
Y1 - 2009
N2 - The ability of ANNs to learn and generalize from examples, and to generate robust solutions, makes them very suitable in a diversity of applications where algorithmic approaches are either unknown or difficult to implement. A major drawback, however, is that the knowledge learned by the network is represented in an exceedingly opaque form, namely, as a list of numerical coefficients. This black-box character of ANNs hinders the possibility of their more wide-spread acceptance. The problem of extracting the knowledge embedded in the ANN in a comprehensible form has been intensively addressed in the literature.
AB - The ability of ANNs to learn and generalize from examples, and to generate robust solutions, makes them very suitable in a diversity of applications where algorithmic approaches are either unknown or difficult to implement. A major drawback, however, is that the knowledge learned by the network is represented in an exceedingly opaque form, namely, as a list of numerical coefficients. This black-box character of ANNs hinders the possibility of their more wide-spread acceptance. The problem of extracting the knowledge embedded in the ANN in a comprehensible form has been intensively addressed in the literature.
UR - http://www.scopus.com/inward/record.url?scp=54849439740&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-88077-6_7
DO - 10.1007/978-3-540-88077-6_7
M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.chapter???
AN - SCOPUS:54849439740
SN - 9783540880769
T3 - Studies in Fuzziness and Soft Computing
SP - 77-81+89-98
BT - Knowledge-Based Neurocomputing
A2 - Kolman, Eyal
A2 - Margaliot, Michael
PB - Springer Berlin Heidelberg
ER -