 |
ML Newsnew | past | best | rss | submit | about |
| Demystifying AlphaGo Zero as AlphaGo GAN (arxiv.org) | |
| 1 point by vog on Dec 1, 2017 | hide | past | pdf | discuss on HN | |
|
In plain words: A close look at AlphaGo Zero's self-play training shows it works like a game where two networks compete and push each other to improve. That built-in rivalry explains why it should train reliably, so its success may not mark a new era of AI. AbstractThe astonishing success of AlphaGo Zero\cite{Silver_AlphaGo} invokes a worldwide discussion of the future of our human society with a mixed mood of hope, anxiousness, excitement and fear. We try to dymystify AlphaGo Zero by a qualitative analysis to indicate that AlphaGo Zero can be understood as a specially structured GAN system which is expected to possess an inherent good convergence property. Thus we deduct the success of AlphaGo Zero may not be a sign of a new generation of AI. | |
| add comment on HN |