XPost: alt.ufo.reports
EXECUTIVE SUMMARY:
- News of a re-connect with Voyager 2 prompted a model to see whether
probe fly-bys of the outer planets is associated with features in
UFO sightings data between (say) 1950 and 2000.
- It seems the best-fit model finds all 4 planets is associated with
features in the sightings data. 2 planets appeared to be associated
with an increase in activity on approach of probe(s) and a decline
after they pass; the other planets are the opposite.
- The fit now explains some of what appeared to be "random walks" in
parts of the UFO sighting data.
- A future model is planned to examine in more detail the "double hit"
from the Voyagers at Jup and Sat that visited months apart in 79 and
then in 80/81.
- It seems everywhere you choose to look there are "signs". Why are
these now-old results not known? Wilful blindness?
Earth to Voyager 2: After a Year in the Darkness, We Can Talk to You Again
The New York Times, 12 Feb 2021 10:43Z
In the nearly 44 years since NASA launched Voyager 2, the spacecraft has
gone beyond the frontiers of human exploration by visiting Uranus, Neptune
and, ...
It will be 44y in mid Aug.
But it got me to thinking if all (if you mean 2) these probes
converging on Jupiter and Saturn in 79 and 80 might have caused some
kind of reaction if there are any "alien bases" out around there.
So we have to build a mathematical model and see whether the sightings
really seem to be affected by some effect of probes getting closer and
closer and closer and then getting further and further away and then
getting closer and closer and ... well, you get the idea.
So the basic element of this model will look like a_i(t_i-d)^r.
I.e. there is a "target date" (t_i) when a probe was due to run past
one of the outer planets, there is the current date (d). And there
are a couple parameters (a_i, r) that need to be estimated to best-fit
the UFO sightings data.
To start with I'll only include the first dates for each of Jup, Sat,
Ur and Nep and gloss over the fact both Voyager 1 and Voyager 2 passed
by Jup and Sat within months so really those 2 components of the model
should be weighted a bit more heavily than Ur and Nep where only V2
visited. But maybe the fitting procedure will end up making the a_i
for Jup and Sat bigger than for the other 2. Like you, I can hardly
wait to see.
I rigged up a simple fitter/optimizer using a genetic search for the
a_i and r on top of a basic time series regression with the overall
aim to maximize the R2 of the final model.
I also tried the "net present threat" type of function with terms like
a_i(1+r)^(|t_i-d|) but these turned out to be a slightly worse fit
than the first cut (above). If seems Our Friends are not ready to
reveal their time-discount rate to us at this point! :)
After a few sec of chugging on what remains of poor old "baby"
() it spat out:
foo.4.108: COEFFICIENT OF DETERMINATION (R SQUARED) = 0.38669
Which is baby speak for "the best set of parameters I can find
explains about 39% of the month-to-month UFO sightings between 1950
and 2000".
The "4.108" is the best value of the "r" parameter and the a_i are
given by (drum-roll please ... ):
LOG TRANSFORM ENABLED
REWEIGHTED LEAST SQUARES BASED ON THE LMS
VARIABLE COEFFICIENT STAND. ERROR T - VALUE P - VALUE
date -0.33735 0.18547 -1.81888 0.06944
t1jup 0.00036 0.00019 1.87776 0.06091
t2sat -0.00072 0.00039 -1.87182 0.06173
t3ura 0.00057 0.00030 1.87929 0.06070
t4nep -0.00021 0.00011 -1.89274 0.05889
CONSTANT 671.52863 367.98557 1.82488 0.06853
WEIGHTED SUM OF SQUARES = 332.94553
DEGREES OF FREEDOM = 583
SCALE ESTIMATE = 0.75570
COEFFICIENT OF DETERMINATION (R SQUARED) = 0.44870
THE F-VALUE = 94.899 (WITH 5 AND 583 DF) P - VALUE = 0.00000
THERE ARE 589 POINTS WITH NON-ZERO WEIGHT.
AVERAGE WEIGHT = 0.97035
Which indicates a slew of interesting things. It seems the sign of the
a_i shows 2 planets seemed to affect UFO sightings in exactly the
opposite way from the other 2, and surprisingly none of the 4 was
"neutral". The procedure *could* have found UFO sightings did not
depend in any way where the probes were with respect to any or all of
the planets on the Grand Tour(s).
It seems Saturn and Neptune were related to sightings across the US
going down as the Voyager(s) approached and then up again after they
passed.
OTOH Jup and Ur as associated with the opposite effect -- as the
probes approached Jupiter sightings across the US/world (this is based
as usual on the NUFORC data which is mostly the US, some Canada, and
comparatively little from the UK, France or the rest of the world)
increased and then decreased after the probes passed.
Contrary to expectation on my part the "first contact" at Jupiter
showed a reaction that was about "half" the fly-bys at Saturn.
Maybe Saturn is "more important" to Our Friends than the others, just
going by the magnitude of the coefficient. The abs values for the a_i
in desc order are Saturn, Uranus, Jupiter and Neptune.
Apart from the effect related to the Voyager probes there is a
constant sightings rate of about 672 per month but that tended (over
the period ~1950-2000) to decrease about 0.3 sightings per month each
year. A tiny effect, but the stats program is "93% sure" it is there
(the P-val on the "date" coeff).
Overall, the s/w says there is almost 0 chance (the p-val on the F
statistic) the Voyager probes had 0 effect on UFO activity/sightings.
The model, above, tracks around 45% of the sighting data (remember
this was pre 2006 where NUFORC changed to a web report form and
subsequently receiving what seems like a lot of junk from trolls and
unhappy incels) and only about 3% of the data 1950-2000 falls more
than 1 sd away from the model.
The plot is at and shows the model in dark
green and the fuzzy cloud of UFO observations in light green.
It seems the otherwise unusual bends in the sightings data are not
totally unexplained.
I'll throw the model building to the AIs to mull over and see whether
they can find a more convincing fit against the full details of
planetary positions and the double fly-bys of Jup and Sat. Can we see
Our Friends hurrying to sweep dust under the rug, let their belts out
again, then hurry to clean up again a few months later at each location?
--
Thought-detection: AI has infiltrated our last bastion of privacy
VentureBeat, 13 Feb 2021 15:09Z
Our thoughts are private - or at least they were. New breakthroughs in
neuroscience and artificial intelligence are changing that assumption, while
at the same ...
[continued in next message]
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
