An Algorithm That Can Predict Human Behavior Better Than Humans (mit.edu)

Quartz describes an MIT study with the surprising conclusion that at least in some circumstances, an algorithm can not only sift numbers faster than humans (after all, that's what computers are best at), but also discern relevant factors within a complex data set more accurately and more quickly than can teams of humans. In a competition involving 905 human teams, a system called the Data Science Machine, designed by MIT master's student Max Kanter and his advisor, Kalyan Veeramachaneni, beat most of the humans for accuracy and speed in three tests of predictive power, including one about "whether a student would drop out during the next ten days, based on student interactions with resources on an online course." Teams might have looked at how late students turned in their problem sets, or whether they spent any time looking at lecture notes. But instead, MIT News reports, the two most important indicators turned out to be how far ahead of a deadline the student began working on their problem set, and how much time the student spent on the course website. ... The Data Science Machine performed well in this competition. It was also successful in two other competitions, one in which participants had to predict whether a crowd-funded project would be considered “exciting” and another if a customer would become a repeat buyer.

Comments

Try predicting violent behavior.

[Ungrounded Lightning]

It might be interesting to try it at predicting future violent behavior of individuals.

Those who study such behavior have come up with the aphorism "The only effective predictor of future violent behavior is past violent behavior." Let's see if the

Shrinks who try to make predictions about individuals come out WORSE than chance - which implies that there may be SOME prediction possible - but the current paradigms have it backward.

This, by the way, is ONE of the reasons the pro-gun crowd pooh-poohs mental health tests for gun ownership or purchase. Another is the observation that people with mental illnesses are, on the average, far LESS likely to cause harm to others than the average of the population. (They may harm THEMSELVES, but suicide rates don't change if guns aren't available: Instead the suicidal switch to less effective and usually more painful means, averaging more tries before they succeed.)

Re:Try predicting violent behavior.

[SuperGus]

The Marshmallow Test has been used to try and assess innate levels of self-control in young kids, and I believe it strongly correlates to things like academic success. https://en.wikipedia.org/wiki/...

My take

[Artem S. Tashkinov]

"MIT News reports, the two most important indicators turned out to be how far ahead of a deadline the student began working on their problem set, and how much time the student spent on the course website. "

IOW, the computer program used a more applicable approach/pertinent data and had humans had used the same algorithm/data, they would have reached the same or better results.

Overblown headline

[binarstu]

The headline of the Quartz article and the Slashdot summary, "An algorithm can predict human behavior better than humans", is, not surprisingly, hugely overblown.

What these researchers actually did was develop a system for automatically taking a massive data set with a huge number of variables, identifying the subset of variables or new combinations of variables that are most likely to be useful for predicting a particular response, and then formulating a predictive model. (This is an extremely simplified summary.) That is really cool, but to present it as some sort of general "algorithm for predicting human behavior" is silly. It's no more an algorithm for predicting human behavior than are automated statistical methods for building a predictive model from a massive dataset.

Re:Overblown headline

[Spy Handler]

Yeah, pretty much. Computer sorts through a massive data set, compiled and fed to it in a convenient format by... humans.

I would be more impressed if they gave the computer eyes and it scanned people walking around and was able to spot some behavior that humans missed. But of course machine vision being what it is currently, computer might not even be able to tell a human apart from a squirrel.

Re:Overblown headline

[Lennie]

If it worked so well as the headline claimed then they wouldn't be writing articles about it, but using it to buy/sell stocks and options. ;-)

No surprise

[gweihir]

First, most humans are not that smart, but do not know that. It is called the "Dunning-Kruger" effect and it is well-established. Apparently at least the OP is unaware of it, possibly making him a subject of the Dunning-Kruger effect.

Second, a majority of humans do not use what smarts they have effectively, but rather do decide "emotionally" when it comes to important decisions or understanding important situations. That obviously works rather badly, just look at what politicians get voted into office, or what life-choices people make. The problem here is that the whole "emotional decision" apparatus is a rather primitive left-over from caveman-times that cannot handle even situations of moderate complexity well. The second problem is that most humans never find that out, as the skill for self-reflection is also rather scarce and hence cannot actively compensate.

So give an arbitrary group of people an analysis or decision problem that somehow "touches" them (like asking students to predict whether other students fail at being students), and suddenly most of them turn into morons (or rather do not stop being morons in many cases), and even a simplistic statistic predictor does a lot better than they do.

Re:Lenovo trackpad ate my homework again. B-b

[gweihir]

There is one: Membership in the biological group "Homo Sapiens". Anything more accurate is just trying to stick labels on people to make them "different" and push them out of society. Some people like doing that to elevate their pathetic selves by stomping on people they perceive to be even more pathetic.

Re:Lenovo trackpad ate my homework again. B-b

[gweihir]

In the past, what came out of these efforts war universally horrible. Like people of a certain descendant being regarded as of low quality and better be disposed of (3rd Reich), people of a certain skin color being regarded as dumb or violent (USA, many others), etc.

On the other hand, nothing useful was ever found. The only ethical thing is to stay away from the whole approach. It only feeds racism and similar mind-sets.

Anti-gun nuts lie again!

[Taco Cowboy]

... Actually suicide rates collapse if you take the guns away ...

To all the anti-gun nuts -

While I understand that your life is so boring that you just have to go around making up all kinds of silly lies ... but please, if you need to lie, please tell lies that are believable !

Do you know that South Korea has a suicide rate which is more than double of that of the United States?

Do you know Japan has a suicide rate of more than 60% higher of that of the United States?

Unlike the Americans, average South Koreans and Japanese don't own guns

http://www.washingtonpost.com/...

Here is a list of countries with their respective suicide rate

Re:Anti-gun nuts lie again!

[Sique]

To the pro-gun nuts: There is a difference between the absolute suicide rate and the number of suicides that could have been avoided if private gun ownership was limited.

There are countries with traditionally high suicide rates, and there are countries with lower rates. Even if the means available to everyone to commit suicide are the same, you still find differences in the number of suicides depending on the region. And then there is the observation that suicide rates behave similar to infection rates of a contagious disease. If in a region suicide rates increase, there is a high probability that suicide rates will increase in the immediately neighboring regions.

Beside that, it is also known that suicidal people are belonging to different groups which use different means. People who have used for instance sleeping pills in an attempt to kill themselves will usually not throw themselves out of a window at the next opportunity, but rather try to kill themselves with sleeping pills again. And people who would use a gun to kill themselves will not jump in front of a subway train and vice versa. So yes, taking away guns would not have forced the 12,000 people in the U.S. who kill themselves with a gun each year to try to gas themselves or to die by causing a traffic accident. Instead, most of those 12,000 people won't have tried to kill themselves at all.

Re:Lenovo trackpad ate my homework again. B-b

[dargaud]

connection between date of birth and violent behavior

Let's divide the duration of the year in say, 12 periods, and give it a scientific sounding name, for instance Astrology...