Google masters Go Gibney, Elizabeth
Nature (London),
01/2016, Letnik:
529, Številka:
7587
Journal Article
Recenzirano
Odprti dostop
A computer has beaten a human professional for the first time at Go - an ancient board game that has long been viewed as one of the greatest challenges for artificial intelligence (AI). To interpret ...Go boards and to learn the best possible moves, the AlphaGo program applied deep learning in neural networks - brain-inspired programs in which connections between layers of simulated neurons are strengthened through examples and experience.
JET's results do not change that, but they suggest that a follow-up fusion-reactor project that uses the same technology and fuel mixture - the ambitious US$22-billion ITER, scheduled to begin fusion ...experiments in 2025 - should eventually be able to reach this goal. JET and ITER use magnetic fields to confine plasma, a superheated gas of hydrogen isotopes, in the tokamak. Last year, the US Department of Energy's National Ignition Facility set a different fusion record: it used laser technology to produce the highest recorded fusion power output relative to power in, a value called Q, where 1 would be generating as much power as is put in.
In social studies, for example, researchers have developed machine-learning models that aim to predict when a country is likely to slide into civil war. ...because the team had tested the model's ...predictive power using data from a time period earlier than some of its training set, the algorithm had effectively been allowed to see the future, he says. ...a computer-vision algorithm might learn to recognize a cow by the grassy background in most cow images, so it would fail when it encounters an image of the animal on a mountain or beach.
PDF