%Aigaion2 BibTeX export from Idiap Publications
%Friday 05 December 2025 08:56:21 AM

@TECHREPORT{Mayo97a,
                      author = {Mayoraz, Eddy},
                    projects = {Idiap},
                       title = {On the Complexity of Recognizing Iterated Differences of Polyhedra},
                        type = {Idiap-RR},
                      number = {Idiap-RR-10-1997},
                        year = {1997},
                 institution = {IDIAP},
                        note = {Published in the Proceedings of ICANN'97},
                    abstract = {The iterated difference of polyhedra $V = P_1 \backslash ( P_2 \backslash (... P_k ) ... )$ has been proposed independently in [Zwie-Aart-Wess92] and [Shon93] as a sufficient condition for $V$ to be exactly computable by a two-layered neural network. An algorithm checking whether $V$ included in $R^d$ is an iterated difference of polyhedra is proposed in [Zwie-Aart-Wess92]. However, this algorithm is not practically usable because it has a high computational complexity and it was only conjectured to stop with a negative answer when applied to a region which is not an iterated difference of polyhedra. This paper sheds some light on the nature of iterated difference of polyhedra. The outcomes are\,: (i) an algorithm which always stops after a small number of iterations, (ii) sufficient conditions for this algorithm to be polynomial and (iii) the proof that an iterated difference of polyhedra can be exactly computed by a two-layered neural network using only essential hyperplanes.},
                         pdf = {https://publications.idiap.ch/attachments/reports/1997/rr97-10.pdf},
                  postscript = {ftp://ftp.idiap.ch/pub/reports/1997/rr97-10.ps.gz},
ipdmembership={learning},
}