Belief Updating and AutoUpdating

When probabilistic inference is done using a Bayes net, it is called belief updating.  New beliefs (posterior probabilities) are found for each of the nodes to reflect the findings currently entered in the net.

Auto-Updating:  Normally Netica does belief updating quickly and automatically after compiling, and every time a finding is entered.  However, it can be slow on large nets, so you may want to enter several findings at once before updating (you also may want to use the optimizing compiler).  In that case, toggle off Network Automatic Updating, enter the findings, and each time you want to update the beliefs choose Network Update, or click the image\UpdateArrowTOOL.gif toolbar button (they will be dimmed if belief updating is not currently required, or if the net is not compiled).

Nothing Happens:  If you enter findings into a net, and no belief updating occurs, perhaps it is because auto-updating is off (the state of auto-updating is saved with each net file) or the net is not compiled.

Dimmed Belief Display:  If auto-updating is turned off, and you enter a new finding, or if no findings are entered but belief updating was never done, then the belief-bars or meters will not display valid results, and so they are drawn dimmed or in gray rather than in solid black.  The reason they are not removed completely is that you may want to see what the beliefs were at the point of the last belief updating.