Title :
Convergent design of a piecewise linear neural network
Author :
Chandrasekaran, Hema ; Manry, Michael T.
Author_Institution :
Dept. of Electr. Eng., Texas Univ., Arlington, TX, USA
Abstract :
A piecewise linear neural network (PLNN) is discussed which maps N-dimensional input vectors into M-dimensional output vectors. A convergent algorithm for designing the PLNN from training data is described The design algorithm is based on a variation of backtracking algorithm known as the `branch-and-bound´ method. The performance of the PLNN is compared with that of a multilayer perceptron (MLP) of equivalent size. The results show that the PLNN is capable of performing as well as an equivalent MLP
Keywords :
convergence; learning (artificial intelligence); neural nets; piecewise linear techniques; tree searching; MLP; PLNN; backtracking algorithm; branch-and-bound method; convergent design; multidimensional input vectors; multidimensional output vectors; multilayer perceptron; piecewise linear neural network design; Algorithm design and analysis; Clustering algorithms; Convergence; Electronic mail; Multilayer perceptrons; Neural networks; Piecewise linear techniques; Tin; Training data; Vectors;
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5529-6
DOI :
10.1109/IJCNN.1999.831157
