Title :
Minimum length filtering with neural networks
Author :
Ingman, D. ; Merlis, Y.
Author_Institution :
Quality Assurance & Reliability, Technion-Israel Inst. of Technol., Haifa, Israel
Abstract :
The derived time evolution equations of the net are similar to the usual continuous Hopfield network, with the exception of shape of the sigmoidal response of the neuron. This difference is a result of the `minimum length´ smoothing condition. The paper also shows that the task of filtration under the weak continuity assumptions can be performed by hidden layer binary neurons. The procedure is demonstrated by two simulations
Keywords :
filtering and prediction theory; image processing; neural nets; signal processing; hidden layer binary neurons; image processing; minimum length filtering; neural networks; sigmoidal response; signal processing; simulations; time evolution equations; weak continuity; Entropy; Equations; Filtering; Filtration; Neural networks; Neurons; Paper technology; Quality assurance; Temperature; Thermodynamics;
Conference_Titel :
Electrical and Electronics Engineers in Israel, 1991. Proceedings., 17th Convention of
Conference_Location :
Tel Aviv
Print_ISBN :
0-87942-678-0
DOI :
10.1109/EEIS.1991.217699
