On Wed, May 23, 2018 at 1:01 PM, Alexey Potapov <[email protected]> wrote:
> 2018-05-22 21:18 GMT+03:00 Linas Vepstas <[email protected]>: > >> >> >> On Sun, May 20, 2018 at 3:26 AM, Alexey Potapov <[email protected]> >> wrote: >> >>> >>> >>> For me, observational data is sensory data. It doesn't contain concepts, >>> predicates, etc. . If we have an observation that a particular crow is >>> black, ... But there are no purely black crows. It's just an abstraction, >>> which itself should be somehow generalized from raw data. >>> How can we calculate P(crow,black|image)? >>> >> >> Do not assume that a probability is what you actually want. Let me give >> three examples. >> >> In real life, when you see a crow, and it is dark, and you want to talk >> about it, you just say "black crow" as an identifier of the object in the >> scene. You don't pull out your photometer and measure it's darkness at >> 87.68% and a blueish hue of 77%. Why? Because you don't need to do that to >> have a conversation about it's presence, location, movement, etc. You only >> need to evaluate crow-ness and blackness sufficiently to distinguish it >> from all other elements of the scene, and then you can assign P==100% for >> most practical purposes. >> >> In neural nets, the sigma-function is a non-linear component, used to >> boost results towards extremes. whatever sum of weights or evidence or >> whatever it is that you have, as inputs feeding the neural net, you apply >> the non-linear sigma, to try to sharpen everything closer to either 0% or >> 100% -- to discriminate. To increase contrast. This is kind-of the >> "secret" as to why neural nets work, and probabilities don't. >> >> In "integrated information" theory, you work with a large complex network >> of things that are all inter-related, all interconnected. The goal of >> applying the theory is to find those extensions of the net that are highly >> interlinked, interconnected, and then to draw an accurate boundary around >> them. If and when you can perceive that boundary, you can give everything >> inside one name, and everything outside a different name. The names >> assigned are unambiguous, unique, even if the actual boundary is perhaps >> uncertain, even if there is a gradation, a smooth-ish transition from the >> highly-interconnected thing, to the mostly disconnected parts. The act of >> name-tagging is what gives a handle on being able to think about the object >> in symbolic terms. >> >> > Well, I cannot agree with you. I can object to the first two paragraphs at > least. However, I'm not sure if it will be productive, since I have a > feeling that the source of discrepancy between our views is mostly in > different definitions... But if you wish, we can continue... > I think we should continue, since we are nominally working together. One cannot work together effectively if there is miscommunication or misunderstandings. These kind of problems don't melt away or evaporate. If you think the problem is definitional, then what do you see as the differences of definition? I tried to give plausible verbal arguments for why things should be a certain way; I don't think I'm being stupid, and I don't particularly see a flaw, per se, at this level. There are, of course 1001 different but important details that need to be resolved, to make things work. The second point about neural nets, is far far more hand-wavey, so if you don't like that one, ignore it. The third point is perhaps much too vague, requiring much too much effort in giving precise definitions, at this time. It was meant to evoke a certain vision. The first point is perhaps the most concrete, in terms of how one can actually proceed, with acutal code that does actual things. --linas -- cassette tapes - analog TV - film cameras - you -- You received this message because you are subscribed to the Google Groups "opencog" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/opencog. To view this discussion on the web visit https://groups.google.com/d/msgid/opencog/CAHrUA34zC%3DgRepYLzQhc7dWGnWEAUcqECdWKSnvQQXT4R3cjfw%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
