TY  - RPRT
TI  - A penalty-function approach for pruning feedforward neural networks.
AU  - R Setiono
PY  - 1997
DO  - 10.1162/neco.1997.9.1.185
UR  - https://pubmed.ncbi.nlm.nih.gov/9117898/
ID  - 9117898
ER  -