This fluctuating sentiment on artificial neural networks is a bit weird; popularity comes and goes in waves, and many academics appear to be just following the hype. Most of the stuff I learned on ANN's in the 90s and early zeros just works, and now we can see that if one throws huge computational power at it, it even works extremely well! Sure there have been a few new tricks added, but for the most of it the deep learning hype just feels like one big 90s connectionists revival party.
Over the years (when they were out of grace) I've had plenty of good results with ANN's in my Go programs (Steenvreter & Magog), I just never went for the depths that are now feasible with modern hardware (but then again, at least my app runs well on a phone:-)). I am amazed by how far the AlphaGo team was able to push it, but the general approach doesn't come as a surprise. If anything is a surprise to me, it's the things they could leave out, and still have it work so well... Erik On Fri, Jan 29, 2016 at 7:06 AM, Detlef Schmicker <d...@physik.de> wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi Ingo, > > I think you are not alone: When I started computer go 4 years ago I > ask a good friend of mine, who did his PhD in Neural Networks back in > the 90s, if I have any chance to use them instead of pattern matching > and he said, they will probably not generalize in a good way :) > > I think the big size of the nets make a qualitative difference, > therefore our intuition is misleading... > > > Congrats to the AlphaGo team, > > Detlef > > Am 29.01.2016 um 02:14 schrieb "Ingo Althöfer": > > Hi Simon, > > > > do your remember my silly remarks in an email discussion almost a > > year ago? > > > > You had written: > >>> So, yes, with all the exciting work in DCNN, it is very > >>> tempting to also do DCNN. But I am not sure if we should do > >>> so. > > > > And my silly reply had been: > >> I think that DCNN is somehow in a dreamdancing appartment. My > >> opinion: We might mention it in our proposal, but not as a > >> central topic. > > > > > > In my mathematical life I have been wrong with my intuition only a > > few times. This DCNN topic was the worst case so far... > > > > Greetings from the bottom, Ingo. > > > > > > > > Gesendet: Donnerstag, 28. Januar 2016 um 16:41 Uhr Von: "Lucas, > > Simon M" <s...@essex.ac.uk> An: "computer-go@computer-go.org" > > <computer-go@computer-go.org> Betreff: Re: [Computer-go] Mastering > > the Game of Go with Deep Neural Networks and Tree Search > > > > Indeed – Congratulations to Google DeepMind! > > > > It’s truly an immense achievement. I’m struggling to think of > > other examples of reasonably mature and strongly contested AI > > challenges where a new system has made such a huge improvement > > over existing systems – and I’m still struggling … > > > > Simon Lucas _______________________________________________ > > Computer-go mailing list Computer-go@computer-go.org > > http://computer-go.org/mailman/listinfo/computer-go > > > >
_______________________________________________ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
