Construction of Artificial Neural Networks for Pattern Recognition Using a Successive Geometric Segmentation Method

This work presents a new method for artificial neural networks construction and training based on successive geometric segmentation (SGSM). This methodology is capable of generating both the network topology and neurons weights without parameters’ specification. The SGSM groups the data of each class into hyper-boxes (HBs) aligned in accordance with the largest axis of its points’ distribution. If the HBs are linearly separable, then a separating hyperplane may be identified resulting in a neuron. If it is not, then a segmentation technique is applied to divide the data into smaller classes for new HBs. For each subdivision, new neurons are added to the network. The tests show a rapid method with a high success rate.

[1]  Manfred M. Fischer,et al.  A Rough Set Approach for the Discovery of Classification Rules in Interval-Valued Information Systems , 2008, Int. J. Approx. Reason..

[2]  Shigeo Abe.  Support Vector Machines for Pattern Classification , 2010, Advances in Pattern Recognition.

[3]  Wenping Wang,et al.  Efficient collision detection using a dual OBB-sphere bounding volume hierarchy , 2010, Comput. Aided Des..

[4]  Thomas Jackson,et al.  Neural Computing - An Introduction , 1990 .

[5]  Glenn Fung,et al.  Multicategory Proximal Support Vector Machine Classifiers , 2005, Machine Learning.

[6]  Jiawei Han,et al.  Data Mining: Concepts and Techniques , 2000 .

[7]  Camillo Gentile,et al.  An improved Voronoi-diagram-based neural net for pattern classification , 2001, IEEE Trans. Neural Networks.

[8]  Jinde Cao,et al.  Robust State Estimation for Uncertain Neural Networks With Time-Varying Delay , 2008, IEEE Transactions on Neural Networks.

[9]  Reshma Khemchandani,et al.  Twin Support Vector Machines for Pattern Classification , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  Claudio Gentile,et al.  Incremental Algorithms for Hierarchical Classification , 2004, J. Mach. Learn. Res..

[11]  Vitor Hugo Ferreira,et al.  Input space to neural network based load forecasters , 2008 .

[12]  Robert I. Damper,et al.  On neural-network implementations of k-nearest neighbor pattern classifiers , 1997 .

[13]  Chih-Jen Lin,et al.  A comparison of methods for multiclass support vector machines , 2002, IEEE Trans. Neural Networks.

[14]  Lutz Hamel,et al.  Knowledge Discovery with Support Vector Machines , 2009 .

[15]  Doug L. James,et al.  BD-tree: output-sensitive collision detection for reduced deformable models , 2004, SIGGRAPH 2004.

[16]  Dinesh Manocha,et al.  OBBTree: a hierarchical structure for rapid interference detection , 1996, SIGGRAPH.

[17]  Dinesh Manocha,et al.  Collision queries using oriented bounding boxes , 2000 .

[18]  N. Bose,et al.  Neural network design using Voronoi diagrams , 1993, IEEE Trans. Neural Networks.