Off Topic Dark Matter and other Astronomy information.

it did, purely because I take you seriously when you post on science matters
My mistake

 

Artificial intelligence called in to tackle LHC data deluge
Algorithms could aid discovery at Large Hadron Collider, but raise transparency concerns.

• Davide Castelvecchi

01 December 2015
GENEVA, SWITZERLAND

Article tools
Rights & Permissions

please log in to view this image

CERN

Particle collisions at the Large Hadron Collider produce huge amounts of data, which algorithms are well placed to process.

The next generation of particle-collider experiments will feature some of the world’s most advanced thinking machines, if links now being forged between particle physicists and artificial intelligence (AI) researchers take off. Such machines could make discoveries with little human input — a prospect that makes some physicists queasy.

Driven by an eagerness to make discoveries and the knowledge that they will be hit with unmanageable volumes of data in ten years’ time, physicists who work on the Large Hadron Collider (LHC), near Geneva, Switzerland, are enlisting the help of AI experts.

please log in to view this image

LHC 2.0: A new view of the Universe

On 9–13 November, leading lights from both communities attended a workshop — the first of its kind — at which they discussed how advanced AI techniques could speed discoveries at the LHC. Particle physicists have “realized that they cannot do it alone”, says Cécile Germain, a computer scientist at the University of Paris South in Orsay, who spoke at the workshop at CERN, the particle-physics lab that hosts the LHC.

Computer scientists are responding in droves. Last year, Germain helped to organize a competition to write programs that could ‘discover’ traces of the Higgs boson in a set of simulated data; it attracted submissions from more than 1,700 teams.

Particle physics is already no stranger to AI. In particular, when ATLAS and CMS, the LHC’s two largest experiments, discovered the Higgs boson in 2012, they did so in part using machine learning— a form of AI that ‘trains’ algorithms to recognize patterns in data. The algorithms were primed using simulations of the debris from particle collisions, and learned to spot the patterns produced by the decay of rare Higgs particles among millions of more mundane events. They were then set to work on the real thing.

please log in to view this image

Computer science: The learning machines

But in the near future, the experiments will need to get smarter at collecting their data, not just processing it. CMS and ATLAS each currently produces hundreds of millions of collisions per second, and uses quick and dirty criteria to ignore all but 1 in 1,000 events. Upgrades scheduled for 2025 mean that the number of collisions will grow 20-fold, and that the detectors will have to use more sophisticated methods to choose what they keep, says CMS physicist María Spiropulu of the California Institute of Technology in Pasadena, who helped to organize the CERN workshop. “We’re going into the unknown,” she says.

Inspiration could come from another LHC experiment, LHCb, which is dedicated to studying subtle asymmetries between particles and their antimatter counterparts. In preparation for the second, higher-energy run of the LHC, which began in April, the LHCb team programmed its detector to use machine learning to decide which data to keep.

LHCb is sensitive to tiny variations in temperature and pressure, so which data are interesting at any one time changes throughout the experiment — something that machine learning can adapt to in real time. “No one has done this before,” says Vladimir Gligorov, an LHCb physicist at CERN who led the AI project.

please log in to view this image

LHC signal hints at cracks in physics' standard model

Particle-physics experiments usually take months to recalibrate after an upgrade, says Gligorov. But within two weeks of the energy upgrade, the detector had ‘rediscovered’ a particle called the J/Ψ meson — first found in 1974 by two separate US experiments, and later deemed worthy of a Nobel prize.

In the coming years, CMS and ATLAS are likely to follow in LHCb’s footsteps, say Spiropulu and others, and will make the detector algorithms do more work in real time. “That will revolutionize how we do data analysis,” says Spiropulu.

An increased reliance on AI decision-making will present new challenges. Unlike LHCb, which focuses mostly on finding known particles so they can be studied in detail, ATLAS and CMS are designed to discover new particles. The idea of throwing away data that could in principle contain huge discoveries, using criteria arrived at by algorithms in a non-transparent way, causes anxiety for many physicists, says Germain. Researchers will want to understand how the algorithms work and to ensure they are based on physics principles, she says. “It’s a nightmare for them.”

Proponents of the approach will also have to convince their colleagues to abandon tried-and-tested techniques, Gligorov says. “These are huge collaborations, so to get a new method approved, it takes the age of the Universe.” LHCb has about 1,000 members; ATLAS and CMS have some 3,000 each.

please log in to view this image

Inside Google's DeepMind artificial intelligence lab

Despite these challenges, the most hotly discussed issue at the workshop was whether and how particle physics should make use of even more sophisticated AI, in the form of a technique called deep learning. Basic machine-learning algorithms are trained with sample data such as images, and ‘told’ what each picture shows — a house versus a cat, say. But in deep learning, used by software such as Google Translate and Apple’s voice-recognition system Siri, the computer typically receives no such supervision, and finds ways to categorize objects on its own.

Although they emphasized that they would not be comfortable handing over this level of control to an algorithm, several speakers at the CERN workshop discussed how deep learning could be applied to physics. Pierre Baldi, an AI researcher at the University of California, Irvine who has applied machine learning to various branches of science, described how he and his collaborators have done research suggesting that a deep-learning technique known as dark knowledge might aid — fittingly — in the search for dark matter.

Deep learning could even lead to the discovery of particles that no theorist has yet predicted, says CMS member Maurizio Pierini, a CERN staff physicist who co-hosted the workshop. “It could be an insurance policy, just in case the theorist who made the right prediction isn’t born yet.”

 

My fact based rebuttal to this was posted weeks before this pseudosciencedaily article

please log in to view this image

4500 joules of UV is 11 on the 1 - 11 risk scale. As can be seen the equator gets 7500 joules in places.

UV kills phytoplankton in the ocean by reacting with "Titanium dioxide nanoparticles" in ocean water (A biotech and nano industry miracle).
There is pollution no one talks about.. phytoplankton is the base of the ocean food chain, if it dies off so does most of the food chain in the oceans.

So without even getting into cancer via UV, there is a much bigger fundamental problem that is being totaly and utterly ignored.

please log in to view this image

UV is up all over the place. The nothern hemisphere where Ozone is lowest, is there we are seeing the biggest temperature changes on the planet.

 

I've moved on from global warming

 

Good read that, have been doing a fair bit of research on AI and computing over the past couple of years, relates loosely to my work.
The AI we are talking about for the LHC is not really AI. It's a command and control system to handle the work load because obviously there is so much data being produced that it is slowing Cern down. I doubt this AI will have a chat with you about the weather or the merits of Vorderman's arse bounding up oand down on you, but it will be able to possibly build models itself, and do a ****load of leg work physicists normally have to do, and pretty much run most of the experiment, so in that context it is more like an operating system than AI tbh.

Quantum computing and AI are not really as advanced as they all let on truth be told, for me part of the issue is they are trying to make hardware neurons, and failing badly, given we still actually know relatively **** all about neuron capability and function.

It certainly makes much sense for them to essentially automate the systems as much as possible or else they need to increase the size of their operation, one or the other. I'd say the latter is far more expensive than the former. IT has been automating for decades. Cern projects are obviously very different from standard IT, but very complicated automation is still automation.

I find it weird that the article says "machine making the discovery", the "machine" is nothing but a hammer, tools dont make discoveries.

 

Read Shadows of the Mind by Professor Sir Roger Penrose. Let me know when you've finished it

 

"A search for the missing science of consciousness." I am intrigued, I might even pick it up rather than read an ecopy.

The human consciousness is nothing but an illusion, it is the cooreration of different parts of the brain. Brain damage patients who just become different people without ever recognising their old self shows that there is no conscious entity. Split brain patients can't even register one half of your face or realise that they cant.

Any advance in this area imo is NOT the current crap of trying to make hardware neurons, our problems are always the same, our processing is serial, base2, we have to increase the pathways and increase throughput. Base2 will never give us ai.

AI that is more like AI will imo involve parts of the human brain hooked up to hardware and the hardware can use the parts, that will be possible before we ever come close to recreating neurons. We still dont know the majority of what the **** neurons do and can do.


Cheers for the book tip, he wrote another The Emperor's new mind. I think I'll pick them both up.

 

I haven't read the other book. It's a tough read, Sis. You can't skim it. It's his theory as to how neurons function at the quantum level. There's a lot of Turing in there and Kurt Godel, so a lot of logic. Certain passages I had to read 15 times or more before I got a basic grasp. It's worth the effort, though - whether you agree or not. It'll keep you away from climate change for a bit

 

At the quantum level you say. Intrigued further, am interested in quantum physics, mainly in terms of finally disproving the limit of the speed of light.

I'll take it for what it is, though, we kind of need to figure out how neurons work before we start talking about neurons at the quantum level. That's going all quantum activist a bit early for me

 

Read the wiki piece on the book. It provides a decent overview

 

I suppose at least Sisu doesn't worry that the LHC will create a black hole to destroy the Earth, because black holes are a #fraud

So there's at least one #conspiracy he doesn't believe in

 

That knee-jerk **** made me laugh from the very beginning

 

will do. I have a knack for skimming over things when it gets a bit deep I usually mark out stuff for further "reading" when I come across stuff I just don't get. As long as it aint full of equations

 

Don't skim. It's much more demanding than your usual 'layman's' science book. but it's well worth it.

 

Adults are talking shhh. Children should be seen not heard

Oh and how is scientists allegedly possibly accidentally creating a black hole (which the whole mainstream media ran with back in the day), how is that< a conspiracy theory?

You are obviously having some sort of break down, coming on here and shouting "look at me me me me", totally what this ^^ post is about

 

If it gets into mathematics, I'm lost, by skim I mean not get bogged down in some thing that might take me weeks to learn, not putting down the book for a week for that reason

 

It's more pure logic than maths

 

Anyone know any computational molecular biologists who want to work in Oxford?

 

It's 3 hours long but try and listen to some of this stuff re. human history etc.............insane stuff

 

I tune into Rogan a bit, 3 hours, will have to watch in parts.