Statistical Methods into Neural Network Applications

Integration of Statistical Methods into Neural Network Applications 

---

Chemometrics provides a number of methods to obtain an insight into data sets, and to extract relevant information from them. The typical chemometric strategy consists of the following steps (a) collection of data for known cases (b) generation of a mathematical model which is usually based on multivariate statistics or neural networks (c) interpretation of the model parameters in terms of the underlying chemistry (d) application of the model to new cases. 

---