Part of Advances in Neural Information Processing Systems 4 (NIPS 1991)
Martin Röscheisen, Reimar Hofmann, Volker Tresp
In a Bayesian framework, we give a principled account of how domain(cid:173) specific prior knowledge such as imperfect analytic domain theories can be optimally incorporated into networks of locally-tuned units: by choosing a specific architecture and by applying a specific training regimen. Our method proved successful in overcoming the data deficiency problem in a large-scale application to devise a neural control for a hot line rolling mill. It achieves in this application significantly higher accuracy than optimally-tuned standard algorithms such as sigmoidal backpropagation, and outperforms the state-of-the-art solution.