
For FullText PDF, please login, if you are a member of IEICE,
or go to Pay Per View on menu list, if you are a nonmember of IEICE.

CombNETIII with Nonlinear Gating Network and Its Application in LargeScale Classification Problems
Mauricio KUGLER Susumu KUROYANAGI Anto Satriyo NUGROHO Akira IWATA
Publication
IEICE TRANSACTIONS on Information and Systems
Vol.E91D
No.2
pp.286295 Publication Date: 2008/02/01 Online ISSN: 17451361
DOI: 10.1093/ietisy/e91d.2.286 Print ISSN: 09168532 Type of Manuscript: PAPER Category: Pattern Recognition Keyword: largescale classification problems, support vector machines, gating networks, divideandconquer,
Full Text: PDF(532.3KB)>>
Summary:
Modern applications of pattern recognition generate very large amounts of data, which require large computational effort to process. However, the majority of the methods intended for largescale problems aim to merely adapt standard classification methods without considering if those algorithms are appropriated for largescale problems. CombNETII was one of the first methods specifically proposed for such kind of a task. Recently, an extension of this model, named CombNETIII, was proposed. The main modifications over the previous model was the substitution of the expert networks by Support Vectors Machines (SVM) and the development of a general probabilistic framework. Although the previous model's performance and flexibility were improved, the low accuracy of the gating network was still compromising CombNETIII's classification results. In addition, due to the use of SVM based experts, the computational complexity is higher than CombNETII. This paper proposes a new twolayered gating network structure that reduces the compromise between number of clusters and accuracy, increasing the model's performance with only a small complexity increase. This highaccuracy gating network also enables the removal the low confidence expert networks from the decoding procedure. This, in addition to a new faster strategy for calculating multiclass SVM outputs significantly reduced the computational complexity. Experimental results of problems with large number of categories show that the proposed model outperforms the original CombNETIII, while presenting a computational complexity more than one order of magnitude smaller. Moreover, when applied to a database with a large number of samples, it outperformed all compared methods, confirming the proposed model's flexibility.

