Title :
Speculative experiment with neural networks on separation and scoring in financial applications
Author_Institution :
Citibank, New York, NY, USA
Abstract :
The difference between separation and scoring with neural networks is analyzed. This study was motivated by financial application of neural networks. A separation corollary is obtained: the better the separation in a BNN (backpropagation neural network) trained with a binary training file, the less the results may be used for scoring
Keywords :
finance; neural nets; BNN; backpropagation; binary training file; financial applications; neural networks; scoring; separation; Backpropagation; History; Intelligent networks; Logistics; Neural networks; Testing; Transfer functions;
Conference_Titel :
Neural Networks, 1991. 1991 IEEE International Joint Conference on
Print_ISBN :
0-7803-0227-3
DOI :
10.1109/IJCNN.1991.170494
