Re: [agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-07 Thread James Bowery
On Tue, Jul 7, 2020 at 2:31 PM Matt Mahoney wrote: > Why bother with a CIC training and test set? Compression evaluates every > bit as a test given the previous bits as training. Even if the compression > algorithm doesn't explicitly predict bits, it is equivalent to one that > does by the chain

Re: [agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-07 Thread Matt Mahoney
Why bother with a CIC training and test set? Compression evaluates every bit as a test given the previous bits as training. Even if the compression algorithm doesn't explicitly predict bits, it is equivalent to one that does by the chain rule. The probability of a string is equal to the product of

Re: [agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-07 Thread Ben Goertzel
COIN CON ;) On Tue, Jul 7, 2020 at 9:22 AM James Bowery wrote: > > A slight modification to avoid idioms like "the HIV virus": > COINC = COIN Criterion > COIN = COmpression Information criteriON > > Otherwise it would be good memetics if its psychological appeal vs memetic > drift reaches the

Re: [agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-07 Thread James Bowery
I'm sure the Hollywood meme machine isn't worried about such quibbles as the "HIV virus" idiom populating its marquees. On Tue, Jul 7, 2020 at 1:43 AM Ben Goertzel wrote: > On Mon, Jul 6, 2020 at 11:41 PM Ben Goertzel wrote: > > > > The COIN Criterion ... sounds like money, it's got to be

Re: [agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-07 Thread James Bowery
A slight modification to avoid idioms like "the HIV virus": COINC = COIN Criterion COIN = COmpression Information criteriON Otherwise it would be good memetics if its psychological appeal vs memetic drift reaches the selective regime in time to achieve fixation against the psychological appeal of

Re: [agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-07 Thread Ben Goertzel
On Mon, Jul 6, 2020 at 11:41 PM Ben Goertzel wrote: > > The COIN Criterion ... sounds like money, it's got to be good... Maybe we can fund the competition by making a Hollywood-style thriller about some Bitcoin criminals... lots of potentials here... > > On Mon, Jul 6, 2020 at 9:13 PM James

[agi] Experimental Testing of CIC (the Compression Information Criterion)

2020-07-06 Thread James Bowery
On Fri, Jul 3, 2020 at 6:53 PM Ben Goertzel wrote: > ...Under what conditions is it the case that, for prediction based on a > dataset using realistically limited resources, the smallest of the > available programs that precisely predicts the training data actually gives > the best predictions