Hi, First an introduction: I am an engineer and did my phd in an application of reinforcement learning to speech technology. I have no formal background in...
This question, "predict the outcome of toss N+1, given tosses 1...N" is a classic Bayesian problem -- it was the topic of Thomas Bayes's original essay, and...
Hi all, I'd like to return to a topic which was discussed recently - the issue of certainty vs knowledge. What I say below will probably be obvious to many, ...
This is discussed in some detail in both Chapter 3-4 and Chapter 18 of PTLOS. ... Yes, that's exactly right. The models M are parameterized by a parameter t,...
I have another question about coin tossing as discussed by Jaynes. He has a section on "Cheating at Coin & Die Tossing". He described a method (using moment...
Thanks David and Kevin. I already have David's book on my computer - and after browsing through it some more it has moved up on my reading list. ... I have not...
I have an issue with Jaynes' discussion in 3.8.1 (p.73-75). Here Jaynes takes the position that the term "random" is meaningless, and goes on to criticise the...
... No, I believe the idea is to throw it so that the axis of rotation is at an angle to the perpendicular to the face of the coin. Something like this, where...
... You're right. I remembered wrong. The detailed discussion is in Chapter 18. ... Because probabilities are a description of one's internal state of mind ...
... The only reasonable and clearcut definition of "random" that I have ever encountered, avoiding the curse of circular definitions, has no direct connection...
Dear Kevin and list members, It is worth noting that the definition based on Kolmogorov complexity is language dependent. That is, how random a finite string...
I have been trained to talk about estimating probabilities. But as Jaynes and Cox pointed out, probabilities are assigned and calculated, relative frequencies...
... I understand that, but the whole point of probability theory is still to enable us to make statements about the real world, which can be verified ...
... This is a different meaning of the word random - it refers to a bit string, but not to individual observations (by this definition a single bit is _always_...
... Even if we decide that the word "calculated" should only be used to refer to the "correct" (JR) calculation, it should still be allowable to say "The...
... (using ... strongly ... Has anyone ever gotten this to work? I think I've partially figured out how to throw the coin, but I don't have a clear idea on how...
... I would think that assigned would be *more* appropriate the less sure you are of your method of prediction. ... You assign the distribution, to make some...
I'm using the list through the web site. Included messages are not wrapping well, so that when I reply to a message, that message is wrapped like ... then jump...
... The point of "probability theory as logic" is explained in the first two chapters. It is to reason consistently and quantatively in the face of ...
Hi Niko, Yes, I agree with what you are saying and I do not think that it contradicts what I am saying. Perhaps I need to describe the sort of experiment I...
... Real coin. As for whether it can be done, I think I at least understand enough on how to throw the coin to get the conditions right - now it is a matter of...
I have the same problem, so I rely on daily digests which are sent to my email address. It's one of the options under the "Edit My Membership" link. I suggest...
... Yes, but the Kolmogorov complexity according to languages L1 and L2 never differ by more than C, where C is a constant that depends on L1 and L2 but not on...
Hi Kevin & others. I scanned Chapoter 18 for the first time this weekend. It seems to disagree with some of the things said in "12.4.3 Unknown probability for...
This weekend I went back and read a part of Chapter 13: 13.7 through 13.9 that I had skipped previously. I was very pleasantly surprised with what I found...
... Yes, but now you just have fixed two languages instead of one. To me this definition of "random" makes sense only if you have fixed some finite set of...
I thought the list members might find this bit of mail I received interesting, as an example of what happens when you stop messing around with ad hoc methods...
There is a feature missing from this product (if it wants to claim to be "Bayesian"): It presents a decision instead of a posterior probability. I think it's...
