The estimate slope of a machine learning prediction function may be shorted routed as x=y to the origin. At x>y or x<y there is a proportional cost/time adjustment variable on short routing on the negative y time axis domain. A predictor algorithm
intercepts the origin & a flop back slopes from a preferred outcome at (O,>0) intercepting at the -x, y<o coordinate and the bottomless draw pot bounce of x=0, y<0 until it hits your clone and you don’t get an identifiable scar. With a nominal cost
obstacle at x = 1 at minimal attentive threshold a prediction slope at time y = 1 will be at cost x<0
https://youtu.be/W3XivHFg9WM
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)