New Crime-Predicting Algorithm Borrows From Apollo Space Mission Tech (digitaltrends.com) 127
Researchers from Georgia Tech and the UK's University of Surrey have developed a new predictive policing algorithm that aims to better manage police resources and gain an upper hand in the war on crime. It reportedly uses technology that's been previously used in weather forecasting and the Apollo space missions. Digital Trends reports: The new algorithm built on previous work carried out by researchers from the University of California and police forces in both the U.S. and U.K. Their 2015 research showed how a predictive policing algorithm could accurately predict between 1.4 and 2.2 times more urban crime than specialist crime analysts. By making recommendations about where to patrol, the algorithm led to a 7.4 percent reduction in crime. However, while effective, this approach has also been criticized due to concerns about possible racial profiling and the underreporting of crime. The new algorithm has so far been demonstrated on a data set of more than 1,000 violent gang crimes in Los Angeles carried out between 1999 and 2002. Early conclusions suggest that the upgraded predictive tool could prove superior for coping with the constantly fluctuating world of real-time crime prediction. The researchers published their paper in the journal Computational Statistics & Data Analysis.
Re: (Score:1)
It's proven wrong hourly, which is why it can't be relied upon. Sure you can follow a statistical trend and maybe find something - or not. What you can't do is say that trend is actually as good as predicting discrete reality, because it isn't. All it can do is extrapolative pattern matching, which sure sometimes will bear fruit, depending - the selection modality implemented is the bias. It's not a human's bias, it's a collection of such biases made algorithmic. Don't be stupid and think it's somehow
Re: This makes no sense (Score:2)
This AI is just a straw man, the police know where to go but cant because they get accused of racial profiling, now they can just say " I was patrolling "area X " because the AI told me to....
Re: (Score:1)
Re: (Score:2)
This isn't AI, but just using statistics calculations for their work.
The problem is most people don't understand statistics. Many who say they do, really don't. Those who do understand it may not be good at it. Many come up with numbers and call it statistics.
Statistical Mathematics hasn't changed much (Well not in comparison with computational improvement) So the algorithm used in the Apallo Space mission which may had taken a few days to run, and give out a chart of numbers. Can now in a split second ru
Police and Rich Fat Old Republicans (Score:1, Troll)
Here's your AI:
Most crimes are caused by corrupt cops or Republicans who run insurance companies, telephone companies, or branded hotels.
Go arrest them and you'll save the rest of us a lot of time and money, and not have to claim 1960s software has anything to do with AI, crime prediction, a Tom Cruise movie (Minority Report), or the price of tea in China (staying stable, but higher for us being saddled with tariffs because of orange-head.)
It's not like you can PREDICT crime. All you can do is statisticall
If they can predict it 3 or 4 days in advance (Score:5, Insightful)
> there's nothing to be done UNTIL the crime is COMMITTED.
If they could predict which block was likely to have a spike in crime 3 or 4 days in advance, that would be enough time to finish their donuts an drive a patrol car over there, potentially preventing the crime. If the presence of the patrol car didn't determine the criminal, the cops could actually catch a bad guy in the act for once, rather than taking a report the next day. I know, that's funny. They'll still just be taking reports after the fact.
Re: (Score:2)
Sue for false advertising. LAPD has their motto "To Protect and Serve" on their cars. Seems to me that a reasonable person would have an expectation of protection.
Re: (Score:3)
Who kills the most people, the police or criminals?
Who hurts more people physically, emotionally and/or economically, the police or criminals?
I think you would get both of those wrong given the putrid bile above but the facts are out there if you'd be interested.
Re: (Score:2)
Who kills the most people, the police or criminals?
False dichotomy. Police commit crimes at roughly the same rate as the general population. Many of them are "the" criminals.
Re: (Score:2)
citation please.
Re: (Score:2)
https://news.vice.com/en_us/ar... [vice.com]
https://en.wikipedia.org/wiki/... [wikipedia.org]
Autocorrect: Deter the criminal, not determine (Score:2)
Autocorrect / predictive text turned "deter the criminal" to "determine the criminal".
It also occurs to me my post was quite negative, and not without good reason, but I should also acknowledge I've dealt with a couple of very good cops. One in particular in Bryan, Texas went above and beyond when my wife was in crisis. On the other hand, 20 miles away in Caldwell, Texas, a different cop wasn't good at all. He shouldn't be a cop.
Re: (Score:2)
The problem is that predicting crime is a self fulfilling prophecy.
The chief of police has limited resources, so they look at their map and see that there is 10% more crime in district A. So they send some extra cops there, and because there are more cops in the are they make more arrests. Now it looks like there is 30% more crime in district A, but the chief spins it as a crackdown with more arrests and bad guys taken off the street.
In reality district A has less crime that district B, but people in distri
Re: If they can predict it 3 or 4 days in advance (Score:2)
Or... Because there is a larger police presence, the criminals move somewhere else temporarily and crime crime rate in area A drops by all appearances....
Re: (Score:2)
That's usually what happens, there is a "crackdown", lots of arrests, then the few criminals who were there move somewhere else and it seems to have worked. At best all it did was displace some crime out of an already low crime area.
A related effect is "return to mean" which is common with speed cameras. There are a few accidents in one place one year, so they put a camera in. The next year the number of accidents falls back to the previous average, so they declare the camera a great success, even though it
Re: (Score:1)
There is this other thing - if the patrol car prevents the crime from occurring in the first place then how is algorithm coping with that? How are algorithm users going to cope with that? I mean all the money spent on y2k bug were wasted because hardly any problems have been seen. The same happened to financial regulation that has been removed which apparently helped the 2008 meltdown to occur. Examples are numerous. Then another q. - if I know a guy or group of guys are up for no good can I do anything - t
Re: (Score:2)
... I mean all the money spent on y2k bug were wasted because hardly any problems have been seen. ...
That was precisely what all the money was spent for, so we would not see so many of the (approaching) problems ! 8-)
Re: (Score:1)
> there's nothing to be done UNTIL the crime is COMMITTED.
If they could predict which block was likely to have a spike in crime 3 or 4 days in advance, that would be enough time to finish their donuts an drive a patrol car over there, potentially preventing the crime. If the presence of the patrol car didn't determine the criminal, the cops could actually catch a bad guy in the act for once, rather than taking a report the next day. I know, that's funny. They'll still just be taking reports after the fact.
I believe the libertarians here would consider this illegal state intervention in the individual's right to commit a crime. Or something.
Re: (Score:2)
What exactly does this have to do with a crime predicting algorithm?
Re: (Score:2)
society never wants to address the root causes of crime, because root causes may make us uncomfortable.
Re:Police and Rich Fat Old Republicans (Score:4, Informative)
Constitutional democracy
Pedantic Jerk here, but we don't live in a Constitutional democracy, we live in a Constitutional Republic. This is a common issue, I understand, but the distinction is important, even if it is ignored. See Electoral College for a prime example of the difference.
The alternative is democracy, which is the last step towards tyranny, as the majority will always vote to oppress the minority.
Re:Police and Rich Fat Old Republicans (Score:4, Insightful)
Look, far be it from me to step on your Pedantic Jerk routine, because I know how much it means to you, but you are wrong.
We live in a democratic republic. Some decisions are made by direct democracy and others are made via democratically-elected representatives. The modifier, "Constitutional" is really pretty meaningless in this context. It just means we have a constitution. That alone does not define or describe our actual governmental system.
Yes, it is a republic. Yes, it is democratic. Yes, you're a Pedantic Jerk, but yes, you're wrong.
This would go against the intent of the founding fathers, who believed the minority should be oppressing the majority. That's why they structured the Constitution so that it takes power out of the hands of people and puts it into the hands of wealthy elite. Unfortunately, they didn't anticipate that in 2018 the "wealthy elite" would end up being a babbling and degenerate prevaricator who hates reading and gets all his information about the world from Fox & Friends.
Re: (Score:1)
I don't care what kind of a government we have, as long as it results in SJWs dancing at the end of a rope.
Re: (Score:2)
What are the *practical* differences between your republic and any other constitutional representative democracy?
Re: (Score:2)
There are no national elections or referendums on direct issues and legislation?
In a Democracy there would be. In a Republic, that stuff is left to the representation of the people.
In most of the world's representative democracies there are no (or very few) referendums.
Re:Police and Rich Fat Old Republicans (Score:4, Insightful)
The alternative is democracy, which is the last step towards tyranny, as the majority will always vote to oppress the minority.
This is rubbish.There are plenty of democracies in Western Europe, and none of them are closer to tyranny than the USA.
Thanks to a voting system that allows more than two parties to flourish, voters organize in smaller blocks and it's pretty rare for one party to have a 50%+ share of the votes. This means government is done by coalitions of two or more parties. There is no single majority block any more, and any oppression of minorities becomes pretty difficult. Politicians also know that vilifying their opponents is counterproductive because one day they may have to form a coalition with the party they're vilifying now. That means party relations are a lot less toxic than in the USA, and coalition governments spend less time dismantling the work of the previous coalition than is common in the US.
Due to the two-party system in the US, it is closer to tyranny (using your definition of 'the majority oppressing the minority') than the West-European democracies.
Re: (Score:3)
Any first-past-the-post system like the US and UK have is likely to produce two party politics. That in turn limits the choice of president to one or two candidates backed by the two big parties. Plus, the whole thing is dominated by money, meaning politicians are effectively all corrupt.
Maybe being a politician should be like being a monk, where you have to take a vow of poverty.
US voting history - collaborations (Score:2)
Politicians also know that vilifying their opponents is counterproductive because one day they may have to form a coalition with the party they're vilifying now. That means party relations are a lot less toxic than in the USA, and coalition governments spend less time dismantling the work of the previous coalition than is common in the US.
I've also read somewhere that in the US, they used to have a time when the president and vice were the two topmost voted, no matter which party they come from (unlike nowadays, where each party sets up a candidate-president, and a candidate-vice, and the come together in the same package, depending on which top party was voted).
I'm laughing trying to imagine an alternate reality where the US kept those law around, and Trump and Clinton were forced to work together (one being the vice president). Trolling th
Re: (Score:2)
This is rubbish.There are plenty of democracies in Western Europe, and none of them are closer to tyranny than the USA.
Tell that to Tommy Robinson who was secretly jailed for protesting outside a courthouse.
Citation
https://www.washingtonpost.com... [washingtonpost.com]
Re: (Score:2)
OK, one democracy in Western Europe is as close to tyranny as the US is. The UK, like the US has a two-party system. My point stands.
Re: (Score:3)
I'll give you a fine example of Tyranny of the majority, in a purely democratic legal vote. Gay Marriage.
The majority, in every single vote on a statewide referendum was voted against gay marriage, or in the affirmative for marriage to be defined as a man and woman only. Even in highly red states like California. The majority in every case voted one way.
The only reason we have gay marriage now, this day, is not because of the majority vote, but rather because of a court that said it violated the Constitutio
Re:Police and Rich Fat Old Republicans (Score:4, Insightful)
there's nothing to be done UNTIL the crime is COMMITTED.
Nonsense. Law enforcement works best when it is focused on PREVENTING crimes, rather than reacting to crimes.
Using stupid phrases like "the war on crime" is not helpful. That is the exact opposite of the mentality we should have.
Historically, by far the most cost effective way to reduce crime has been to improve prenatal and early childhood nutrition and remove environmental pollutants, especially heavy metals such as lead, cadmium, and mercury. Today, black children in America have twice the average blood lead levels as white children, and prison inmates have three times the average. There is a lot more we can do at very low cost.
Re: Police and Rich Fat Old Republicans (Score:4, Informative)
Other effective ways of reducing crime are not having the rich live next door to the poor
Not true. Mixed income neighborhoods do not have more crime. The worst crime is in areas with concentrated poverty, where the poor prey on the poor.
Re: (Score:1)
The worst crime is in areas with concentrated poverty, where the poor prey on the poor.
Back in college, I helped a friend of mine move in to a poor neighborhood. It was really worrisome to be there. Every house had bars on every window and door. Broken junk laying in yards and driveway. And also a group of ~20-30 guys just standing in the street watching us the whole time we moved him in.
Re: (Score:1)
where the poor prey on the poor
That's the worst.
Maybe it doesn't get reported to the police as much.
That way it would not show up on crime heat maps as much.
Re: (Score:2)
They're already working on redistributing the wealth.
Crime does not redistribute wealth. It is mostly poor people stealing from other poor people. If a thief steals your wallet, he may get $20 in cash, but cause you hundreds of dollars in time and expense to get new credit cards, apply for a replacement DL, etc. So net wealth is destroyed. High crime neighborhoods have high prices (to make up for the robberies), and low investment in job creation. Poor people spend far more of their disposable income on locks, window bars, and other security measures.
Re: (Score:2)
Historically, by far the most cost effective way to reduce crime has been to improve prenatal and early childhood nutrition and remove environmental pollutants, especially heavy metals such as lead, cadmium, and mercury. Today, black children in America have twice the average blood lead levels as white children, and prison inmates have three times the average.
Citation needed.
Re: (Score:2)
I'll make you a deal. We take all the "corrupt cops or Republicans who run insurance companies, telephone companies, or branded hotels" and place them in a lock-down facility.
If the crime rate decreases significantly during that period, you win and we keep them there.
If it doesn't, we shoot you in the head and exterminate your entire family, to ensure your gene pool doesn't continue.
Seems a good bet if you're reasonably sure. Deal?
Re: (Score:2)
Or, you know, do both. Work for social cohesion, and in the meantime lock up murderers and rapists. If you have an infection, by all means take something for the fever and pain even while you take antibiotics for the root cause.
Re: (Score:1)
What exactly is "social cohesion"? Is that the kind of police work they currently have in Britain where they arrest people for saying things the government deems as hate speech? Do you achieve "social cohesion' by using the power of the state to enforce same-think? I've read that book and I don't want a live in a society like that.
Re: War On Crime is part of Wars On Stuff series (Score:1)
No. Social cohesion means that people don't blame each other for their common problems.
In that context hate speech is just ridiculous. Arresting people for it would give it undeserved credence, and is an effective means of preventing social coherence.
Opportunity Costs (Score:1)
But what are the opportunity costs of people having to bear the burden of living under a predictive model? Until these can be outlined in totality can one truly make a determination whether or not to implement such measures?
What is the cost of a person complying with new regulations. Who is going to pay these costs, and how do these costs show up on the balance sheet of the proposal?
Re: (Score:1)
But what are the opportunity costs of people having to bear the burden of living under a predictive model? Until these can be outlined in totality can one truly make a determination whether or not to implement such measures?
Of course. See "prohibition", which the instigators truly thought would be a good idea. Turns out, not so. Then they had to retreat a bit... and proceeded to criminalise other drugs instead. Most of the drug-associated crime is really only there because of the laws prohibiting the substances. They could've regulated instead.
But the point is, of course you can implement measures before you have a good idea of what they'll do. Politicians do it all the time. It's one reason why they have such a bad reputation
Re: (Score:3)
But what are the opportunity costs of people having to bear the burden of living under a predictive model?
If done right, there is very little negative. The cops are already driving around patrolling and if we assume that a person is less likely to commit a crime if they see a cop drive by then by telling the cops where to patrol reduces crimes, reduces arrests, and everyone benefits.
Most crimes whether it is vandalism, rape, speeding, or even theft are crimes of opportunities. You aren't going to try to steal a person's phone if that person is watching you. You aren't going to speed if you see a cop. By str
Re: (Score:3)
Assuming that solely for the sake of argument, what guarantees do we have it will be "done right"? Will this stay the case? How do you propose to ensure and transparently show that this is indeed the case? And so on. Do tell.
This isn't predicting people, this is predicting locations. Even if it got so good that it knows there is going to be a robbery at the convenient store on 9th street at 9pm (which it can't), the cop isn't going to arrest someone until they see the crime being committed. This is similar to a tip being called in saying "I think there is going to be a gang fight at the pier at 9pm". The cop doesn't go down there and start arresting people. If there is a crowd, they might disperse the crowd and tell them t
Re: (Score:3)
If there is a crowd, they might disperse the crowd and tell them to go home but they still aren't arresting people until an actual crime is committed.
Nonsense. Cops arrest people who haven't committed a crime all the time [criminaljusticelaw.org].
Re: (Score:3)
Those are all good questions. But those questions are all one sided. Without asking the question of the other half of the equation, "what is the cost of crime", we're never going to fix anything correctly.
Keep in mind, the communities that most affected by crime are ALSO the communities that have the same profile as the criminals in that community. IF we are okay with subjecting them to high crime out of fear of offending people for "Racial Profiling", that is pretty damn racist IMHO. It is racism of low ex
Re: (Score:3)
I have heard people on left complain that those in certain communities are more likely to be targeted and arrested and convicted.
That is an excuse, right there. People who live in high crime areas, commit higher numbers of crimes. It is "cause and effect". How can one stop crime, if you're so fucking sensitive about the race of the perpetrators that you make excuses as to why society can't arrest the criminals because of their race, ignoring the victims are also of the same race.
The "targeted" part of your comment is racially charged, and part of exactly what I was suggesting, it is the worst form of racism, because it preve
Re: Opportunity Costs (Score:1)
But what are the opportunity costs of people having to bear the burden of living under a predictive model?
Exactly none.
You are free to ignore weather reports and it won't cost you anything. That way you won't benefit from the predictions either, but that opportunity cost is not incurred by the prediction.
Mathematics, is there anything it can't do? (Score:3)
This is a novel application the Kalman filter ("technology that's been previously used in weather forecasting and the Apollo space missions") which I would expect most electrical engineers to have experience with if they've ever done any sensor applications (this is the second step to processing sensor data if simple averaging doesn't work).
What TFA and, it's associated paywalled, journal article shows that regardless of what is the situation, chances are there is a mathematical approach to dealing with it. The problem is finding somebody who can look at the problem and present it in a way that different mathematical approaches can be used.
This doesn't mean that the Kalman filter is the *best* approach for this problem, just that it is a novel way of looking at the issue of predicting crime.
Re: Mathematics, is there anything it can't do? (Score:2)
Where does it say that they are using a Kalman filter?
I would rather expect them to use a Runge-Kutta method, but I am not an expert in simulation models.
Anyway, using mathematics to predict crime is not at all new. It is called predictive policing and has shown some success in London for years now. Mind you, it only predicts crime. It won't help prevent it if it isn't used by city planners instead of police.
Re: (Score:1)
Where does it say that they are using a Kalman filter?
From the abstract: "A novel and general Bayesian sequential data assimilation algorithm is developed for joint state-parameter estimation for an inhomogeneous Poisson process by deriving an approximating Poisson-Gamma ‘Kalman’ filter that allows for uncertainty quantification."
So, not a 'real' Kalman filter? But is still doing a joint probability distribution approximation? Not sure what it's really doing, but that's where the Kalman comes in. I also have no idea why it would be better than
Re: (Score:2)
google "kalman filter"
Profiling? (Score:2)
How is it racial profiling? Who's putting race into the prediction models?
It's violent gang related crime profiling.
Re:Profiling? (Score:4, Insightful)
How is it racial profiling? Who's putting race into the prediction models?
Even if race is not explicitly included, it is likely strongly correlated with other inputs, such as zipcodes, neighborhoods, names, prior arrest records, profile of the victim (most crime is intra-racial), etc. Since REALITY is that crime is correlated with race, any "AI" will quickly locate proxies for race in the data.
If we want to eliminate racial bias, it is not effective to modify the inputs. It is better to just modify the outputs by multiplying by a race based fudge factor to make the results politically acceptable.
Re:Profiling? (Score:4, Insightful)
Yeah, better to make the police less effective incase someone's feelings get hurt.
Re: (Score:2)
The algorithm doesn't care about race, only in crimes, so isn't racist unless there are more crime reported by racists due to being racists - and then the racism isn't in the algorithm.
Re: (Score:2)
The algorithm doesn't care about race
In theory, "separate but equal" wasn't racially biased either. But if the output is racist, then then most people would say that the system is racist.
Race is correlated with criminality. Any AI that isn't specifically designed to exclude that factor will find that correlation and use it to make predictions. We, as a society, have decided that is unacceptable because it doesn't treat people as individuals. So do you fudge the inputs or fudge the outputs? Fudging the inputs is difficult, unlikely to work
Re: (Score:2)
Re: (Score:2)
More police, faster to respond. Less time to escape.
Politically the GUI will show exactly what parts of a city and state have huge amounts of crime.
Re: (Score:2)
Are you people really this fucking stupid or is it just convenient to be deliberately obtuse?
Race is getting into the data used by the prediction models because of the many decades of racial profiling used by cops.
If cops routinely charge poor and black kids for possession of dope while routinely letting off white middle-class kids with a warning, then that's going to affect the crime stats that influence the model. Same when they routinely stop and frisk black people but not whites - a percentage of those
Re: (Score:2)
These statistical models work based on crime REPORTING. So you're saying citizens in urban centers are over-reporting crimes and whites are under-reporting crimes? It must be all those white people coming up with phrases like "Snitches get stitches" and "fuck da police"
Re: (Score:2)
It's certainly not banker fraud related crime profiling where it might actually do some good.
they should start the pre-crime unit (Score:2)
wait, isn't that what happened in minority report?
Part of the "war on freedom" series ... (Score:2)
There really is no reason for a "war on crime", except as a pretext to fight freedom and to make some people even richer.
Re: (Score:2)
Move to Somalia and get all the freedom you want.
Re: Part of the "war on freedom" series ... (Score:3)
It's an experiment! (Score:1)
C'mon folks. We're not talking Future Crimes stuff here. They plugged in data, had statistical models run over it, and did some testing to see what happened.
If there's a certain area where there is more reported crime then increasing police presence will help in reducing crime in that area. Of course there will be secondary effects such as underreported crime and criminals moving to other areas that aren't as heavily patrolled.
The first effect - underreported crime - is something that can't be addressed sin
Hmmmm... (Score:1)