Constrained classifier: a novel approach to nonlinear classification

Abbassi, Hadj Ahmed; Monsefi, Reza; Yazdi, Hadi Sadoghi

doi:10.1007/s00521-012-1194-9

Cited by 1 publication

(1 citation statement)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A specific type of local classification is based on the idea of pairwise coupling between positive and negative examples or clusters is conceptually close to the initialization we propose for our LDNN model. These methods typically employ a clustering algorithm, learning classifiers between pairs of positive and negative clusters found by clustering, finally followed by a combination scheme such as voting to integrate the pairwise classifiers into a single decision [38], [39], [40], [41], [42], [43], [44]. The modular network [21] discussed previously also falls into this category.…”

Section: Related Workmentioning

confidence: 99%

Disjunctive normal networks

2016

View full text Add to dashboard Cite

Abstract-Artificial neural networks are powerful pattern classifiers; however, they have been surpassed in accuracy by methods such as support vector machines and random forests that are also easier to use and faster to train. Backpropagation, which is used to train artificial neural networks, suffers from the herd effect problem which leads to long training times and limit classification accuracy. We use the disjunctive normal form and approximate the boolean conjunction operations with products to construct a novel network architecture. The proposed model can be trained by minimizing an error function and it allows an effective and intuitive initialization which solves the herdeffect problem associated with backpropagation. This leads to state-of-the art classification accuracy and fast training times. In addition, our model can be jointly optimized with convolutional features in an unified structure leading to state-of-the-art results on computer vision problems with fast convergence rates. A GPU implementation of LDNN with optional convolutional features is also available

show abstract