Upgrade September 27, 2007
(edited November 7, 2026)
Forecast models based on planetary speeds
In this upgrade I have found a way to work with planetary speeds smarter. Actually, you were able to work with planetary speeds before, using this ULE event:

For example, this is the diagram of Mercury speed:

However, the usage of this type of events for the forecast
is not a good idea. It is much better to forecast with the planetary speeds if
we apply the ideas that come from Fuzzy Logic science. Instead of using the speed
itself, we deal with events like this one: "Mercury is direct and its
speed is the highest" or "Mercury is retrograde, and its speed is
medium".
To create this kind of events in the Neural Network module, follow these steps:

Here chose this tab:

See what events the program creates for the Moon (as an example):

1) 20% of the time ("20" - because the number of grades=5) the speed of the Moon is lower than 12º02min
2) Another 20% of the time the speed of the Moon is between 12º2min and 12º44min
3) Next speed range is 12dº44min and 13º34min
4) 13º34īmin and 14º13min
5) finally, 20% of time the speed of the Moon is higher than 14º13min
The most important issue regarding this way of breaking the Moon's speed is
that they are of equal time intervals. See how these events for graded Moon's
speed look in time:

Starting from this upgrade, you are able to apply these
events to generate the Neural Network projection line using all planets in different
Zodiacs:

The projection line based on graded planetary speed events
appears too steppy because a graded speed event is essentially a Boolean event.
It only indicates whether the planetary speed falls within a specific speed
range, corresponding to a membership function in fuzzy logic.

We recommend smoothing this projection line using this function:
