New algorithm lets machines learn like humans

Scientists have created a new algorithm that captures human-like learning abilities. The algorithm enables computers to recognize and draw simple handwritten characters after exposure to just a few examples.

See Full Article

Their work, summarized in an article published Thursday in the journal Science, represents a major advancement in the field of machine learning.

The new algorithm, called “Bayesian Program Learning,” attempts to mimic the way humans learn new concepts. When humans are exposed to a new concept – say a new symbol or object– they can often recognize and understand the new concept after being exposed to it just a few times.

During a teleconference, U of T assistant computer science professor and article co-author Ruslan Salakhutdinov, gave the example of thumbs-up or high-five gestures that humans can recognize and perform as needed, after limited exposure to them.

While there are now machines and programs that can also learn to recognize symbols and patterns, for example object or speech recognition, these machines often depend on tens or hundreds of examples in order to function accurately.

But the new algorithm can recognize and learn a new concept after a limited number of examples, the authors of the study say. As well, the algorithm can perform more creative tasks beyond simple recognition. These tasks include generating new concepts, breaking down concepts into their respective parts, and understanding the relationship between those respective parts.

Salakhutdinov said the algorithm is a significant advancement in the field of artificial intelligence.

"It has been very difficult to build machines that require as little data as humans when learning a new concept," he said in a statement. "Replicating these abilities is an exciting area of research connecting machine learning, statistics, computer vision, and cognitive science."

Recognizing and reproducing handwritten characters

Using the algorithm, a computer was able to create a copy of handwritten characters, after being shown an example. In this case, the new characters were letters from different alphabets from around the world.

The computer was also able to mimic the way humans draw characters, by using similar pen strokes, stroke order and direction, the researchers said.

Finally, the computer was able to generate a totally new character when shown a set and asked to produce another one to be included in the set.

When the researchers compared the computer-generated results with characters produced by humans, they found them to be "mostly indistinguishable."

Machine learning

Can you tell the difference between humans and machines? Humans and machines were given an image of a novel character (top) and asked to produce new exemplars. The nine-character grids in each pair that were generated by a machine are (by row) B, A; A, B; A, B. (Photo courtesy Brenden Lake)

Machine learning

Can you tell the difference between humans and machines? Humans and machines were given an image of a novel character (top) and asked to produce new exemplars. The nine character grids in each pair that were generated by a machine are (by row) 1, 2; 2, 1; 1, 1. (Photo courtesy Brenden Lake)

Brenden Lake, the study's lead author and a Moore-Sloan Data Science Fellow at New York University, said the algorithm can lead to the development of better learning machines.

"Our results show that by reverse engineering how people think about a problem, we can develop better algorithms," he said in a statement. "Moreover, this work points to promising methods to narrow the gap for other machine learning tasks."

Joshua Tenenbaum, a professor at MIT in the Department of Brain and Cognitive Sciences and the Center for Brains, Minds and Machines and a co-author of the article, echoed the sentiment.

"I've wanted to build models of these remarkable abilities since my own doctoral work in the late nineties," he said in a statement. "We are still far from building machines as smart as a human child, but this is the first time we have had a machine able to learn and use a large class of real-world concepts – even simple visual concepts such as handwritten characters – in ways that are hard to tell apart from humans."

'Bayesian Program Learning'

The algorithm works by representing concepts as simple computer programs.

For example, the letter "A" is represented by computer code, which will generate examples of that letter when the code is run, NYU said in a statement released Thursday.

However, no human programmer is required during the learning process, rather, the algorithm programs itself by building code to produce the letter it sees. And, unlike standard computer programs that produce the same output every time, these special programs produce different outputs every time they’re run, the statement said.

This captures the variation that exists among concepts, such as the natural differences between how two people might draw the letter "A," the statement added.

In a video explaining his work, Lake said that while the algorithm only currently works for handwritten characters, it could eventually be used in other domains.

"The key point is that we need to learn the right form of representation, not just learning from bigger data sets, in order to build more powerful and more human-like learning patterns," he said.



Advertisements

Latest Tech & Science News

  • 'Monster bird' fossil found in New Zealand

    Tech & Science CTV News
    The fossilised bones of an ancient penguin the size of a pro-wrestler have been discovered in New Zealand, scientists said Tuesday, dubbing the creature "monster bird." With an average height of 1.7 metres (5.5 feet) and a weight of 100 kilogrammes (220 pounds), the giant bird is thought to be one of the world's biggest extinct penguin species, easily dwarfing its cuddly-looking modern descendants. Source
  • North American birders flock to N.B. tree after rare bird from Europe spotted

    Tech & Science CTV News
    MIRAMICHI, N.B. -- New Brunswick has a rare visitor from Europe and it's attracting bird lovers from across Canada and the United States. A European mistle thrush arrived in Miramichi on Saturday, and decided to stay. Source
  • Birth of new island could help search for life on Mars: NASA

    Tech & Science CTV News
    NASA researchers are studying the formation of a new Pacific Ocean island in order to find clues for where to search for past life on Mars. Located in the South Pacific nation of Tonga, Hunga Tonga-Hunga Ha'apai began forming after the eruption of an underwater volcano in Dec. Source
  • Fetid attraction: London fatberg to go on museum display

    Tech & Science CTV News
    LONDON -- Part of a monster fatberg that clogged one of London's sewers is destined for fame in a museum. The Museum of London says it will put the only remaining chunk of the 130-metric-ton (143-U.S. Source
  • Telescope to scan mysterious cigar-shaped asteroid for signs of alien technology

    Tech & Science CTV News
    If E.T. has a cellphone, astronomers are hoping to find it. Researchers will use a listening telescope to search for signs of alien technology on ‘Oumuamua, the mysterious, fast-moving, cigar-shaped interstellar object currently speeding through our solar system. Source
  • Pregnant woman wants seat on Tokyo metro: there's an app for that

    Tech & Science CTV News
    Pregnant women struggling to bag a seat on the famously packed Tokyo subway could find their salvation in a new app connecting them with nearby passengers willing to give up their coveted perch. The digital match-making app being trialled this week on the metro aims to overcome two problems especially prevalent in Japan: passengers generally have their nose buried in their phones and talking is strictly frowned upon. Source
  • SpaceX delivery via recycled rocket delayed a day

    Tech & Science CTV News
    CAPE CANAVERAL, Fla. - SpaceX has delayed its latest grocery run for the International Space Station for at least a day. The company now aims to launch its first recycled rocket for NASA on Wednesday. Source
  • Arctic report card: Permafrost thawing faster than before

    Tech & Science CTV News
    NEW ORLEANS - A new report finds permafrost in the Arctic is thawing faster than ever before. The annual report card released Tuesday also finds water is warming and sea ice is melting at the fastest pace in 1,500 years at the top of the world. Source
  • Tokyo airport to be 'scattered' with robots for 2020 Olympics

    Tech & Science CTV News
    Visitors to the 2020 Tokyo Olympics can expect to arrive at an airport "scattered" with robots to help them, an official said Tuesday as he unveiled seven new machines to perform tasks from helping with luggage to language assistance. Source
  • NASA to announce major planet-hunting discovery made possible by AI

    Tech & Science CTV News
    NASA is poised to make a significant announcement involving the search for Earth-like planets, in an effort that has been aided by artificial intelligence designed by Google. NASA and Google have scheduled a news conference to reveal their findings at 1 p.m. Source