4.3. Neural network information

4.3.1. Supervised neural network information

The information window for a supervised neural network shows some information on the network topology, the attribute that were used and the relative importance (weight) on a scale from 0 to 100 of the individual attributes. Be careful to judge these weights, this information tells a lot about individual attributes, but only a little about relationships between attributes. This information can be stored into a report (a simple text file) by pressing the Save Report button. Press Dismiss to go back to the Neural Network Manager.

The information window for unsupervised neural networks shows the essential information on the training: the number of classes and the average match at the moment the training was stopped.

The table of data at the bottom of the information window shows the class centers. In case the neural network was trained on samples alone (i.e. a seismic waveform) you can visualize the class centers by pressing the Display button. This pops up a seismic viewer window with the class centers displayed as wiggle traces.

This information on the neural network can also be saved into a report (a simple text file) by pressing the Save Report button. Press Dismiss to go back to the Neural Network Manager.

4.3.2. Seismic viewer window

The traces are displayed in the 2D Viewer. The scale on the left are the sample index numbers, always starting at 0. Zoom in by clicking the left button in the 2D Viewer or by selecting an area manually with the left mouse-button. The values of each of the wavelets are displayed in the lower left corner of the viewer and change when moving along.

The 2D Viewer has several options;

• zooming in on the display

• zooming out

• move upward

• move left

• move right

• move down

• flip the section

• set the display parameters