Previous Next

*   Neural Networks Module
The Neural Networks module displays classification results from Self-Organizing Maps (SOM), which are a special class of neural networks.
A self-organizing map is a network of neurons arranged in the form of a two-dimensional (2-D) lattice. You can define the size of a lattice or let the application automatically calculate it. During a classification, neurons become selectively activated to various input spectra as a result of a competitive learning process.
The objective of classification analysis using SOM is to find classes of spectra on the map that exhibit common or similar properties. When one or more spectra activate the same neuron, you can assume the spectra belong to a common class. In this case, the spectra should exhibit certain similarities. You can consider spectra that activate neighboring neurons and those neurons that have low Euclidian distance between each other (shown by borderline thickness), to be related.
In neural networks, each mass spectrum must always activate a neuron and this spectrum is shown on the particular neuron. Neurons are displayed as rectangles. Spectra are represented as symbols or numbers, and are placed onto neurons. Because the spectra are located in discrete objects, the interpretation of SOM is relatively easy in contrast to PCA and Fuzzy Clustering where you do not have diffuse clusters. However, when a single spectra activates a larger numbers of neurons, this advantage is lost.
Spectra that activate the same neuron belong to the same classification pattern. All spectra drawn inside a neuron box are equal for classification purposes and their graphical positions within a neuron are irrelevant.
Note  The SOM classification method exhibits one atypical feature to be aware of: Different results are produced for an identical data set if the input data is processed in a different order. This impact of order on results is an inherent feature of neural networks and not a result of faulty algorithms.
For additional information, see Self-Organizing Maps or Fuzzy Clustering.


Previous Next
Related Topics:
  Generating a Neural Networks Module
  Using the Neural Networks Window
  Accessing Spectra from a Neural Networks Window
Copyright 1998 - 2013 HighChem Ltd., Slovakia