Affordable Access

Publisher Website

Analysing complex sensory data by non-linear artificial neural networks

Authors
Publisher
Elsevier Science & Technology
Identifiers
DOI: 10.1016/s0922-3487(96)80028-1
Disciplines
  • Computer Science
  • Design

Abstract

Publisher Summary This chapter analyzes complex sensory data by nonlinear artificial neural networks. The problems associated with prediction in sensory science can be seen as twofold. The first is that data are often nonlinear. Simple transformations such as logarithms may help, but it may be that the nonlinearities are more complex. A second problem is that while good models may be obtained where the variation between samples is large, this is seldom the case where variation between samples is very small. Nonlinear methods like neural networks should be taken into account to detect the nonlinear relations in the data. Nonlinear methods like neural networks should be taken into account to detect the nonlinear relations in the data. The transfer function used in the neural net is designed to detect both linear and nonlinear relations in the data. For analyzing complex sensory data process it is required to use different tools giving corresponding results but with different degrees of accuracy. It is very important for the user of neural nets and principal component regression/partial least squares regression (PCR/PLS) to understand the limitations and pitfalls.

There are no comments yet on this publication. Be the first to share your thoughts.

Statistics

Seen <100 times
0 Comments

More articles like this

Predictive non-linear modeling of complex data by...

on Current Opinion in Biotechnolo... February 2002

Predictive non-linear modeling of complex data by...

on Current Opinion in Biotechnolo... Jan 01, 2002

Non-linear modelling of13C NMR chemical shift data...

on Journal of Molecular Structure Jan 01, 1994

Non-linear modelling of13C NMR chemical shift data...

on Journal of Molecular Structure... Jan 01, 1994
More articles like this..