Title :
Some n-bit parity problems are solvable by feedforward networks with less than n hidden units
Author :
Setiono, Rudy ; Hui, Lucas Chi Kwong
Author_Institution :
Dept. of Inf. Syst. & Comput. Sci., Nat. Univ. of Singapore, Singapore
Abstract :
Starting with two hidden units, we train a simple single hidden layer feedforward neural network to solve the n-bit parity problem. If the network fails to recognize correctly all the input patterns, an additional hidden unit is added to the hidden layer and the network is retrained. This process is repeated until a network that correctly classifies all the input patterns has been constructed. Using a variant of the quasi-Newton methods for training, we have been able to find networks with a single layer containing less than n hidden units that solve the n-bit parity problem for some value of n. This proves the power of combining quasi-Newton method and node incremental approach.
Keywords :
Newton method; feedforward neural nets; learning (artificial intelligence); optimisation; pattern recognition; feedforward neural network; hidden units; n-bit parity problems; node incremental approach; optimisation; pattern recognition; quasi-Newton methods; Computer science; Error correction; Feedforward neural networks; Feedforward systems; Information systems; Network topology; Neural networks; Newton method; Pattern recognition; Transfer functions;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.713918
