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

  • Police sketch created with DNA technology is potentially useless or even misleading, says scientist

    Tech & Science CBC News
    Calgary police hired a U.S. biotech company to create a picture of a woman using only her DNA, but a scientist says the evidence behind the technology to create the image simply isn't there. Source
  • Neanderthals, not humans, created these cave paintings in Spain

    Tech & Science CBC News
    From the murky depths of Spanish caves comes a surprising insight: Neanderthals created art. That's been proposed before, but experts say two new studies finally give convincing evidence that our evolutionary cousins had the brainpower to make artistic works and use symbols. Source
  • Neanderthals, not modern humans, created these cave paintings

    Tech & Science CBC News
    From the murky depths of Spanish caves comes a surprising insight: Neanderthals created art. That's been proposed before, but experts say two new studies finally give convincing evidence that our evolutionary cousins had the brainpower to make artistic works and use symbols. Source
  • Science Says: European art scene began with Neanderthals

    Tech & Science CTV News
    NEW YORK - New discoveries in some Spanish caves give the strongest evidence yet that Neanderthals created art. The key finding was the age of some previously known cave paintings and decorated seashells. European researchers found they were created before our species arrived in Europe. Source
  • Fishing industry has massive global footprint: study

    Tech & Science CTV News
    HALIFAX -- Global fishing efforts are so wide ranging that fleets covered more than 460 million kilometres in 2016 -- a distance equal to going to the moon and back 600 times. That startling revelation is contained in a newly published study in Science that quantifies fishing's global footprint for the first time. Source
  • Twitter bars tactics used by 'bots' to spread false stories

    Tech & Science CBC News
    Twitter Inc. said on Wednesday it would no longer allow people to post identical messages from multiple accounts, cracking down on a tactic that Russian agents and others have allegedly used to make tweets or topics go viral. Source
  • Social media may be pushing more millennials to turn to cosmetic procedures, clinics say

    Tech & Science CBC News
    At age 27, Vanessa Alaumary has already had several cosmetic procedures. She started getting injectables a few years ago and says many of her friends also started in their early 20's. Selfie craze triggers surge in cosmetic procedures Source
  • The other Pyeongchang Games: eSports have taken over South Korea

    Tech & Science CTV News
    While thousands of athletes compete for Olympic glory in Pyeongchang, just blocks away hundreds of other athletes are competing for digital glory in front of computer screens. eSports, basically the sport of playing video games, has blown up in South Korea thanks in part to the country’s fibre-optic network -- said to be the world’s best. Source
  • Feds lift all bovine tuberculosis quarantines from Prairie farms

    Tech & Science CTV News
    EDMONTON -- The federal government says all quarantines imposed on Prairie farms and ranches because of a few cases of bovine tuberculosis have been lifted. Agriculture Minister Lawrence MacAulay says no additional cases of the infectious disease have been detected. Source
  • More than half of federal government scientists still feel muzzled, poll finds

    Tech & Science CBC News
    More than two years after the election of a federal government that says it wants scientists to speak freely, more than half of federal scientists who respond to a new poll say they still don't feel they can. Source