Function approximation is to find the underlying relationship from à given finite input-output data. It has numerous applications such as prediction, pattern recognition, data mining and classification etc. Multilayered feed-forward neural networks (MLFNNs) with the use of back propagation algorithm have been extensively used for the purpose of function approximation recently. Another class of ...

This paper presents the use of back propagation neural network to implement voice recognition. The focus is to identify voice patterns of different people so as to recognize their voices electronically. The signals corresponding to a text phrase of a group of people are recorded in voice files on a computer using sound recording software. The information in these files is converted from time do...

John A. Hertz Nordita Blegdamsvej 17 2100 Copenhagen Denmark We introduce a cost function for learning in feed-forward neural networks which is an explicit function of the internal representation in addition to the weights. The learning problem can then be formulated as two simple perceptrons and a search for internal representations. Back-propagation is recovered as a limit. The frequency of s...

Sandbox experiments with different boundary conditions demonstrate that antiformal stacks result from a forward-breaking thrust sequence. An obstacle blocks forward thrust propagation and transfers the deformation back to the hinterland in a previously formed true duplex. In the hinterland, continued shortening causes faults to merge toward the tectonic transport direction until the older thrus...

The back-propagation algorithm is widely used for learning in artificial neural networks. A challenge in machine learning is to create models that generalize to new data samples not seen in the training data. Recently, a common flaw in several machine learning algorithms was discovered: small perturbations added to the input data lead to consistent misclassification of data samples. Samples tha...

Abstract This document will mainly be of interest to those who are re-implementing or extending “Choosing a Variable to Clamp” [1]. Most of the text is concerned with demonstrating a fairly straightforward isomorphism between forward and reverse-mode automatic differentiation applied to the belief propagation algorithm, in section 3. This shows that “Back-Belief Propagation” of [1] is performin...

a neural network with feed forward topology and back propagation algorithm was used to investigate the effect of composition on mechanical properties in api x65 microalloyed steel (used in manufacturing of large diameter pipes). experimental data was obtained by cutting 100 specimens from pipes manufactured in industrial scale (with similar heats and manufacturing processes). the chemical analy...

common heat transfer fluids such as water, ethylene glycol, and engine oil have limited heat transfer capabilities due to their low heat transfer properties. nanofluids are suspensions of nanoparticles in base fluids, a new challenge for thermal sciences provided by nanotechnology. in this study, we are to optimize and report the effects of various parameters such as the ratio of the thermal co...