[FRIAM] Judea Pearl: Book of Why
Steven A Smith
sasmyth at swcp.com
Sun Apr 19 09:25:32 EDT 2020
Frank -
> Steve,
>
> I'm surprised you didn't find any posts by me in your search for
> "causality" .
Actually I did find you, your voice was in that thread as well, and in
others.
> Usually, when someone says "correlation is not causation" it
> triggers me. In the early 90s/late 80s there were two teams working
> on inference of causal graphs from observational data: Pearl et al at
> UCLA and Glymour et al at Carnegie Mellon. They cooperated and
> developed algorithms based on d-separation which was based on
> conditional independence relations (correlation). Glymour et al's
> book is "Causation, Prediction and Search". I implemented many
> algorithms in Java for that group over a period of about 10 years and
> I was co-author of several papers.
>
> Sorry for the narcissistic reaction.
I am glad you spoke up immediately. If I had something actionable to do
with this right now, I would be consulting your grounded experience, but
also welcome your broader perspective. I saw your references to Glymour
and wish (also) that I had followed that "back then". I think the
first discussions I saw here were as early as 2003? I snagged on the
2013 one because of Doug and Tory's "voices" from the other side of the
veil.
I was lead to Pearl's latest by a colleague who has been encouraging me
to look at the Pandemic Modeling challenge in a somewhat different way
than has been traditional. Essentially a space-time network of causal
implications ( I think) which would be rich in post-hoc elaboration.
The simplest SIR models are challenged with not knowing nearly enough
about the Infected in particular, and even then after they have been
Infectous for days or even weeks. The recent hints (Santa Clara
county?) that infection rates may be 10x or higher than believed (a or
sub or crypto -symptomatic) demonstrate that acutely. My work of late
(other than SimTable) has been in the realm of trying to analyze
ensembles of predictive simulations. This is a logical next step
(forward and backward propogating data and constraints as they are
recorded/discovered/postulated) across space (populations) and time.
More offline maybe?
- Steve
>
> Frank
>
> ---
> Frank C. Wimberly
> 140 Calle Ojo Feliz,
> Santa Fe, NM 87505
> 505 670-9918
> Santa Fe, NM
>
> On Sun, Apr 19, 2020, 6:26 AM Steven A Smith <sasmyth at swcp.com
> <mailto:sasmyth at swcp.com>> wrote:
>
> Glen -
>
> I'm in the midst (early part) of Judea Pearl's "Book of Why".
>
> I had a vague memory of his earlier book: "Causality" having been
> referenced if not discussed on this list. Searching the
> archives, I discovered what I considered to be quite a Pearl (NPI)
> circa 2013. In this long chain, you recommended Pearl's
> "Causality" which I now wish I had followed up on then.
>
> http://friam.471366.n2.nabble.com/beyond-reductionism-twice-td7582273i20.html#a7582308
>
> Among the many gems in the thread were the voices of two of our
> deceased members, Doug Roberts and Tory Hughes. Doug coined one
> of his classic lines about (paraphrase) "being violently
> disinterested in the philosophy of causation" (or complexity or
> agent-based-model-design).
>
> After Nick's recent "violent disinterest in the Cult of Feynman"
> and in particular to any quote that might imply that birds are
> (paraphrase) "not first-class-citizens who would in fact be
> interested in ornithology, if they were given access to it", my
> eyes caught on your own quote (in 2013) of S. Ulam:
>
> "Talking about non-linear mathematics is like talking about
> non-elephant
> zoology." -- Stanislaw Ulam
>
> - Steve (176)
>
>>> Unfortunately, after a couple of attempts to read it, I couldn't understand anything in your post except this part. My previous post was just under 300 words. So, I decided to try to make the next one under that mark as well.
>>>
>>> On 4/18/20 1:22 PM, Steven A Smith wrote:
>>>> From whence (or wherest?) did you get your 300 word target?
>> you might not be alone in that... perhaps it was just gibberish. And likely more than three hundred words of it.
>>
>>
>> .-. .- -. -.. --- -- -..-. -.. --- - ... -..-. .- -. -.. -..-. -.. .- ... .... . ...
>> FRIAM Applied Complexity Group listserv
>> Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam <http://bit.ly/virtualfriam>
>> unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
>> archives: http://friam.471366.n2.nabble.com/
>> FRIAM-COMIC http://friam-comic.blogspot.com/
> .-. .- -. -.. --- -- -..-. -.. --- - ... -..-. .- -. -.. -..-. -..
> .- ... .... . ...
> FRIAM Applied Complexity Group listserv
> Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam
> <http://bit.ly/virtualfriam>
> unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> archives: http://friam.471366.n2.nabble.com/
> FRIAM-COMIC http://friam-comic.blogspot.com/
>
>
> .-. .- -. -.. --- -- -..-. -.. --- - ... -..-. .- -. -.. -..-. -.. .- ... .... . ...
> FRIAM Applied Complexity Group listserv
> Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam
> unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> archives: http://friam.471366.n2.nabble.com/
> FRIAM-COMIC http://friam-comic.blogspot.com/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20200419/f8e6688e/attachment-0001.html>
More information about the Friam
mailing list