TY  - GEN
KW  - neural network 
KW  -  hardware 
KW  -  evolutionary algorithm 
KW  -  classification
TI  - Stepwise Evolutionary Training Strategies for Hardware Neural Networks
Y1  - 2005///
ID  - heidok5529
A1  - Hohmann, Steffen Gunther
N2  - Analog and mixed-signal implementations of artificial neural networks usually lack an exact numerical model due to the unavoidable device variations introduced during manufacturing and the temporal fluctuations in the internal analog signals. Evolutionary algorithms are particularly well suited for the training of such networks since they do not require detailed knowledge of the system to be optimized. In order to make best use of the high network speed, fast and simple training approaches are required. Within the scope of this thesis, a stepwise training approach has been devised that allows for the use of simple evolutionary algorithms to efficiently optimize the synaptic weights of a fast mixed-signal neural network chip. The training strategy is tested on a set of nine well-known classification benchmarks: the breast cancer, diabetes, heart disease, liver disorder, iris plant, wine, glass, E.coli, and yeast data sets. The obtained classification accuracies are shown to be more than competitive to those achieved by software-implemented neural networks and are comparable to the best reported results of other classification algorithms that could be found in literature for these benchmarks. The presented training method is readily suited for a parallel implementation and is fit for use in conjunction with a specialized coprocessor architecture that speeds up evolutionary algorithms by performing the time-consuming genetic operations within a configurable logic. This way, the proposed strategy can fully benefit from the speed of the neural hardware and thus provides efficient means for the training of large networks on the used mixed-signal chip for demanding real-world classification tasks.
UR  - https://archiv.ub.uni-heidelberg.de/volltextserver/5529/
AV  - public
ER  -