Artificial intelligence: Learning machine navigates London Underground
Nature
2016년10월13일
A hybrid learning machine that combines the best features of neural networks and computers is described in a study published in Nature this week.
Conventional computers can process complex forms of data, but require manual programming to perform these tasks. Artificial neural networks have been developed to mimic brain-like learning that can identify patterns in data, but they lack the memory architectures needed for symbolic processing of structured data.
Alex Graves, Greg Wayne, Demis Hassabis and colleagues developed a so-called ‘differentiable neural computer’ (DNC), which comprises a neural network that can learn by example or through trial and error, and an external memory structure similar to random-access memory in a conventional computer. Thus, it can learn like a neural network but process complicated data like a computer.
The study shows the DNC can successfully understand graph structures like family trees or transport networks; for example, planning the best route on the London Underground without prior knowledge of this transport system or solving moving block puzzles with goals described in a symbolic language.
doi: 10.1038/nature20101
리서치 하이라이트
-
6월24일
Sport science: New wearable sensor to measure neck strain may detect potential concussionScientific Reports
-
6월23일
Scientific community: Women credited less than men in scientific paper authorshipNature
-
6월17일
Health technology: New cost-effective smartphone test for middle ear functionCommunications Medicine
-
6월16일
An exercise-inducible molecule that suppresses appetiteNature
-
6월16일
Cancer: Signatures of structural genomic variation in cancerNature
-
6월10일
Animals: Genetic clues to how dogs became man’s best friendsScientific Reports