[tt] Next Big Future - 7 new articles

[Eugen Leitl]

----- Forwarded message from FeedBlitz <feedblitz at mail.feedblitz.com> -----

From: FeedBlitz <feedblitz at mail.feedblitz.com>
Date: Mon, 14 Dec 2009 02:04:43 -0500
To: eugen <eugen at leitl.org>
Subject: Next Big Future - 7 new articles
Reply-To: FeedBlitz <feedblitz at mail.feedblitz.com>
X-Mailer: FeedBlitz

** "Next Big Future" - 7 new articles - http://feedblitz.com/r.asp?l=26702798&f=64651&u=2987253&c=0

- Biomarker Detection of Whole Blood
- Rumor of the Discovery of Particles of Dark Matter
- Superconducting Wire Improving Fast and Will Enable Many Applications
- Argonne scientists discover mechanism behind superinsulation
- Carnival of Space 133 with North Pole Mysteries, astronomy and Future space colonization
- Development Path for Helion Energy for 2010 and 2012
- Google Using Dwave Systems Quantum Computer as a binary classifier of images
- More Recent Articles
- Search Next Big Future

* Biomarker Detection of Whole Blood - http://feedblitz.com/r.asp?l=42980939&f=64651&u=2987253&c=0

Blood is filtered and transferred to nanosensors on a chip, which can detect and measure cancer biomarkers. Credit: Mark Reed/Yale University

A team led by Yale University researchers has used nanosensors to measure cancer biomarkers in whole blood for the first time. They used nanowire sensors to detect and measure concentrations of two specific biomarkers: one for prostate cancer and the other for breast cancer. 

To overcome the challenge of whole blood detection, the researchers developed a novel device that acts as a filter, catching the biomarkers—in this case, antigens specific to prostate and breast cancer—on a chip while washing away the rest of the blood. Creating a buildup of the antigens on the chip allows for detection down to extremely small concentrations, on the order of picograms per milliliter, with 10 percent accuracy. This is the equivalent of being able to detect the concentration of a single grain of salt dissolved in a large swimming pool. 

Until now, detection methods have only been able to determine whether or not a certain biomarker is present in the blood at sufficiently high concentrations for the detection equipment to give reliable estimates of its presence. "This new method is much more precise in reading out concentrations, and is much less dependent on the individual operator's interpretation," Fahmy said. 

The new device could also be used to test for a wide range of biomarkers at the same time, from ovarian cancer to cardiovascular disease, Reed said. "The advantage of this technology is that it takes the same effort to make a million devices as it does to make just one. We've brought the power of modern microelectronics to cancer detection."

Nature Nanotechnology: Label-free biomarker detection from whole blood

Label-free nanosensors can detect disease markers to provide point-of-care diagnosis that is low-cost, rapid, specific and sensitive. However, detecting these biomarkers in physiological fluid samples is difficult because of problems such as biofouling and non-specific binding, and the resulting need to use purified buffers greatly reduces the clinical relevance of these sensors. Here, we overcome this limitation by using distinct components within the sensor to perform purification and detection. A microfluidic purification chip simultaneously captures multiple biomarkers from blood samples and releases them, after washing, into purified buffer for sensing by a silicon nanoribbon detector. This two-stage approach isolates the detector from the complex environment of whole blood, and reduces its minimum required sensitivity by effectively pre-concentrating the biomarkers. We show specific and quantitative detection of two model cancer antigens from a 10 µl sample of whole blood in less than 20 min. This study marks the first use of label-free nanosensors with physiological solutions, positioning this technology for rapid translation to clinical settings.

19 page pdf of supplemental information

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;cancer,biomarkers,medicine,disease,personalized medicine;Biomarker Detection of Whole Blood;3089110 • Related • •

* Rumor of the Discovery of Particles of Dark Matter - http://feedblitz.com/r.asp?l=42980694&f=64651&u=2987253&c=0

The physics blogs are abuzz with rumors that a particle of dark matter has finally been found.

The Cryogenic Dark Matter Search experiment is one of several designed to look for the tell-tale signature of dark matter particles passing through

CDMS is located deep underground in the Soudan mine in Minnesota, to protect it from the hail of cosmic rays that would otherwise wash out any dark matter signal. The gossip mill went into overdrive after a rumour leaked out that the CDMS collaboration has had a paper accepted by the journal Nature. Word is that the paper will appear in the 18 December issue.

CDMS Statement - December 10, 2009 

The CDMS collaboration has completed the analysis of the final CDMS-II runs, which more than doubled the total data from all previous runs combined. The collaboration is working hard to complete the first scientific publication about these new results and plans to submit the manuscript to arXiv.org (http://arXiv.org) before the two primary CDMS talks scheduled for Thursday, December 17, 2009 at Fermilab and at SLAC. Jodi Cooley, the CDMS analysis coordinator and a professor from Southern Methodist University, will present the talk at SLAC at 2 p.m. PST, and Lauren Hsu, a scientist from Fermilab, will present the talk at Fermilab at 4 p.m. CST. A Web cast of Cooleys talk will be available on the CDMS Web site (http://cdms.berkeley.edu/) 

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;controversial,science,speculation,physics;Rumor of the Discovery of Particles of Dark Matter;3089110 • Related • •

* Superconducting Wire Improving Fast and Will Enable Many Applications - http://feedblitz.com/r.asp?l=42974241&f=64651&u=2987253&c=0

In 2006 an 2007, the Department of Energy set up targets for superconducting wire like the following

Performance of HTS wires is expressed in amperes-meters (A-m), where the critical current of 1 cm-wide wire at 77K and self-field is multiplied by the wire length. The A-m value provides a basis for the comparison of different HTS wires, and is an actual indicator of progress in wire development. HTS wires with a current capacity of 1,000 amperes and piece-length of 1,000 meters, i.e., equivalent of 1,000,000 A-m,

will satisfy the requirement of most practical applications.

Demonstrate prototype 100,000 A-m critical current-length for second generation wire by FY12

Demonstrate prototype 500,000 A-m critical current-length for second generation wire by FY14

Demonstrate prototype 800,000 A-m critical current-length for second generation wire by FY15

Demonstrate prototype 1,000,000 A-m critical current-length for second generation wire by FY20

Development of HTS wires and electric power equipment, such as cables, fault current limiters, and transformers

The work seems to be a few years ahead of schedule and should have significant commercial impact in 2014-2018.

2010 Plans at SuperPower Inc

Manufacturing of long-length, high current wires at high throughput

• Bring new pilot MOCVD manufacturing system with nine-track helix

system into full operation and more than double manufacturing capacity

• Increase performance of routine production 2G wires by 50% to 300 A/cm

over lengths of several hundred meters

Improved pinning for enhanced high-field performance

• Develop new techniques for consistent nanocolumnar structures to fabricate thick HTS films MOCVD with high critical currents in high fields

– Achieve a 35% retention of self-field critical current at 77 K and 1 T, in films thicker than 2 μm

• Transition new approaches for nanocolumnar structures within the HTS film for improved consistency of in-field performance of production 2G wires

– Improve critical current of 500+m wires in a field of 0.52 T at 77 K to 150 A/cm from current level of 90 A/cm, with uniformity better than 5%

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;future,energy,technology,superconductor;Superconducting Wire Improving Fast and Will Enable Many Applications;3089110 • Related • •

* Argonne scientists discover mechanism behind superinsulation - http://feedblitz.com/r.asp?l=42972452&f=64651&u=2987253&c=0

An electron microscopy image of titanium nitride, on which the effect of superinsulation was first observed. 

Scientists at the U.S. Department of Energy's Argonne National Laboratory have discovered the microscopic mechanism behind the phenomenon of superinsulation, the ability of certain materials to completely block the flow of electric current at low temperatures. The essence of the mechanism is what the authors termed "multi-stage energy relaxation."

Traditionally, energy dissipation accompanying current flow is viewed as disadvantageous, as it transforms electricity into heat and thus results in power losses. In arrays of tunnel junctions that are the basic building units of modern electronics, this dissipation permits the generation of current. 

Argonne scientist Valerii Vinokour, along with Russian scientists Tatyana Baturina and Nikolai Chtchelkatchev, found that at very low temperatures the energy transfer from tunneling electrons to the thermal environment may occur in several stages. 

“First, the passing electrons lose their energy not directly to the heat bath; they transfer their energy to electron-hole plasma, which they generate themselves,” Vinokour said. “Then this plasma 'cloud' transforms the acquired energy into the heat. Thus, tunneling current is controlled by the properties of this electron-hole cloud.”

As long as the electrons and holes in the plasma cloud are able to move freely, they can serve as a reservoir for energy—but below certain temperatures, electrons and holes become bound into pairs. This does not allow for the transfer of energy from tunneling electrons and impedes the tunneling current, sending the conductivity of the entire system to zero.

“Electron-hole plasma disappears from the game and electrons cannot generate the energy exchange necessary for tunneling,” Vinokour said.

Because the current transfer in thin films and granular systems that exhibit superinsulating behavior relies on electron tunneling, the multistage relaxation explains the origin of the superinsulators. 

Superinsulation is the opposite of superconductivity; instead of a material that has no resistivity, a superinsulator has a near-infinite resistance. Integration of the two materials may allow for the creation of a new class of quantum electronic devices. This discovery may one day allow researchers to create super-sensitive sensors and other electronic devices. 

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;materials,science,united states,physics;Argonne scientists discover mechanism behind superinsulation;3089110 • Related • •

* Carnival of Space 133 with North Pole Mysteries, astronomy and Future space colonization - http://feedblitz.com/r.asp?l=42968121&f=64651&u=2987253&c=0

1. Above is a piece of the 370 megapixel image of 500,000 galaxies.

Phil Plait, the bad astronomer, discusses the huge image just released by the Canada-France-Hawaii Telescope Legacy Survey Deep Field #1, a ginormous mosaic of the night sky. 

It covers a solid square degree of sky — 5 times the area of the full Moon — and tips the scale at a whopping 370 megapixels! It took 5 years and several hundred hours of observing time with the 3.6 meter telescope on top of Mauna Kea to get this massive mosaic. The image itself may look cool and all, but the true power comes when you give in to the dark side you use the interactive zoom feature

2. 

Cat Dynamics covers the Giant Green Spiral over the north pole and also covers the Case for Pluto as a planet.

3. Universe Today also talks about "What was the Norway Spiral?"

4. "A Solid Look at Sail Technologies" for Centauri Dreams:

This reviews an article that ran in a magazine produced by CUNY -- the article covers Greg Matloff's work on solar sails and also discusses Roman Kezerashvili's interesting studies of how a sail moving close to the Sun would be affected by General Relativity.

5. Steve's astrocorner discusses NASA Wise infrared space telescope.

If all goes like the simulations performed at Cal-Tech then dozens of brown dwarfs should be found within 25 light years of Earth. Other items to scan for will be dark asteroids. Dark asteroids present a hazard to Earth in the form of a run in.These asteroids do not reflect light very well and because of this, cannot be picked up in a telescope.There are estimated 100,000 of these Asteroids orbiting undetected. WISE will be able to see these Astro-boulders because they absorb the Sun's heat and then reflect it in the infrared 

6. Alan Boyle's cosmic log at msnbc has firstly

Spaceship debut causes chills

Hundreds gather at a Mojave Desert airport for the unveiling of SpaceShipTwo, which is likely to be the world's first commercial spaceship. Those hundreds flee soon after the unveiling, due to a windstorm that sweeps over the airport.

7. Alan Boyle's cosmic log at msnbc has secondly

From the desert to space

SpaceShipTwo isn't the only game in town: Mojave is also home to other ventures that are targeting the final frontier, and making money as they do it.

8. Alan Boyle's cosmic log at msnbc has lastly

Light show sparks UFO buzz

A spectacular light show visible from northern Norway energizes the UFO crowd, but experts determine that the display was actually caused by a failed Russian ballistic missile.

9. A Really Cool Exoplanet from the astroblogger

Discusses the visual discovery of an almost solar system like gas giant around a near duplicate of our sun by the Subaru telescope.

10. Planetaria discusses the new evidence for past life on Mars

11. This week at One-Minute Astronomer... an exclusive interview with world-renowned astrophotographer Jerry Lodriguss. In this interview, you'll learn the basics of taking a simple but quite lovely photo of the night sky with a digital camera. No telescope required. 

12. NASA Chandra X-ray Observatory blog has "Galaxy Collision Switches on Black Hole"

 

13. Collectspace has design submissions from Astronauts, space workers for NASA's end-of-shuttle patch contest

14. From "A Babe in the Universe" Direct from Johnson Space Center Building 31, where it all happened, we hear the latest on: "Life From Mars"

In 1996 colleagues discovered signs of life on Martian meteorite ALH84001. The latest paper demolishes competing explanations, making the case for life ever stronger.

15. Weirdwarp looks at various planets and moons and their potential for being terraformed.

16. Simostronomy gives some advice on operating an amateur telescope in cold weather

17. Alice's Astroinfo has a gift guide

Looking for something for that little (or big) astronomer in your life? Well, if you need a little help, or feel a like the possibilities are too infinite, I’ve gathered together some of my suggestions here.

18. Cheap Astronomy delivers a podcast on the local stellar neighbourhood of our little corner of the Milky Way.

19. TheSpacewriter visits an online multiwavelength explorer.

I mention this a lot in my public talks in various venues: that the sky we see with our eyes isn’t the sum total of the universe that can be detected. I’ll say it again, another way. When you look at the night sky with your eyes, you’re only seeing the universe through a very small window of emissions.

20. Cumbrian Sky looks at Eddington Crater

I’m the Secretary of Kendal’s astronomical Society, the “Eddington Astronomical Society,” and that it is named after the famous astrophysicist Sir Arthur Eddington, who was born here in Kendal in 1882.

21. Cumbrian Sky discusses the Virgin Galactic Spaceship Two

22. Here at nextbigfuture:

Suborbital tourism is a stepping stone to two hour flights around the world which would then lead to large scale orbital and space travel.

23. Still here at nextbigfuture: 

Historical colonization and details on the age of sail and how it helps understand how to make appropriately scaled plans to colonize space.

* In the age of sail cross ocean colonization, navies faced off with 150 ships on a side and send fleets of 1-11 ships carrying hundreds per ship to explore and colonize and trade with the newly discovered Americas.

* By the late 16th century American silver accounted for one-fifth of Spain's total budget. 

* In the 16th century "perhaps 240,000 Europeans" entered American ports.

 

* If (when) there is human settlement of space and if there were parallels to the scale of the settlement of the Americas, then there would be thousands of spaceships capable of carrying hundreds of people at a time for interplanetary and later interstellar travel. The interplanetary capability (out to the Oort comet cloud) would be something like the ships traveling and trading around the mediterranean.

24. And finally, at nextbigfuture this week

Current aircraft can be made comfortable while keeping passenger density up. This relates to future spaceships passenger seating configurations

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;space,nasa,Mars,carnival of space,astronomy;Carnival of Space 133 with North Pole Mysteries, astronomy and Future space colonization;3089110 • Related • •

* Development Path for Helion Energy for 2010 and 2012 - http://feedblitz.com/r.asp?l=42943786&f=64651&u=2987253&c=0

Helion Energy Researchers proposing two intermediate systems that last one could be useful for fusion/fission hybrid on the way to full fusion energy

Development of a High Fluence Fusion Neutron Source and Component Test Facility Based on the Magneto-kinetic Compression of FRCs by John Slough who is working on developing a Fusion Reactor

This was discussed on the Energy From Thorium forum It was noted that the fusion/fission hybrid would not be as affordable or as simple as the liquid flouride thorium reactor (LFTR). However, the fusion transmuter hybrid might be available as early as 2012 for $30-40 million in development while a LFTR would take longer. It was noted that it would take 16.5 years for the fusion transmuter to produce Uranium 235 needed. However, if each module only costs $20-30 million then $300 million for ten modules would cut the production time down to 1.65 years. Nuclear reprocessing plants currently are very expensive. $20 billion for the Japanese facility (Rokkasho). Five hundred fusion transmuter modules would cost about $15 billion (and the price could go down with factory mass production). 

In the latest Helion Energy presentation they discuss 50 fusion engines being able to burn or transmute the entire stockpile of nuclear waste in 20 years. The Fusion engine is an upgrade of the 2012 system.

The chart is saying that a full commercial fusion engine would cost less than 100 million and is ballpark estimated at about $30-70 million

50 times $100 million - $5 billion is far cheaper than waste repositories or reprocessing.

If the 2012 unit is not the fully ready thing, based on the timelines it could be 2-4 years more to get to a useful commercial system.

Say 2016.

The $2.5 million proposal looks like ARPA-E size or government stimulus fundable thing. Plus U of Washington gets enough in its regular physics budget to pay for it.

Still faster than development of a LFTR.

Potentially faster and cheaper than building a repository. (especially with regulatory and political issues)

Just the transmutation part even without the dedicated fission reactors would be worthwhile. I think they could transmute fuel for some of the regular reactors now to use

The Lawrence Livermore (LIFE hybrid) and other fusion fission hybrid systems were all talking about one billion dollar plus development and 2020-2030 as timeframes for when the main development would be occuring at the earliest.

Motivation for CTF (Component Test Facility) based on the FRC (Field Reversed configuration)

Criteria for Component Test Facility:

(1) Provide an environment close to the fusion reactor

(a) On the smallest physical scale (cost and timeliness)

(b) With the simplest configuration (cost and ease of use)

(2) It should be capable of evaluating the full tritium fuel cycle.

(3) It should allow for easy diverter access for evaluation of a range of materials evaluation

Magneto-kinetic Compression of FRC plasmoids

Fusion Power Density scales as β2B4

(1) The FRC has the highest of all fusion plasmas ( ~ 0.8-0.9)

(2) Compression and burn occurs in a simple linear geometry at highest

possible B consistent with pulsed solenoidal coil (Bz ~15-20 T)

(3) D-T fusion neutron generation provides more realistic test for

materials and tritium breeding

(4) Diverter outside blanket and remote from burn chamber

The pulsed FRC based CTF:

(1) Reduces by orders of magnitude the scale and complexity involved in a CTF based on a spallation source, ST or tokamak

(2) Provides for a vastly lower cost, risk and a much shorter timescale for implementation

(3) Can address both material exposure issues in blanket as well as diverters.

(4) Can address crucial Tritium fueling concerns –

(I) Inventory, (II) Production, (III) Processing, and (IV) Recovery

All are without resolution

All represent potential show-stoppers for DT fusion.

(5) Can be further developed to contribute to energy production in the near term

(I) fissile/fusile breeder

(II) Waste transmutation/burner

UPDATE

Energy from Thorium forum has two follow up comments that are relevant for this article

From Lars: The claim for 50 fusion engines to consume the entire US stockpile in 20 years is based on a Sandia report 

In this you will find that in addition to the fusion neutron source supplying a small percentage of makeup neutrons one must also:

separate the actinides from the spent fuel

develop a lead cooled, fluid reactor

develop a first wall material that can survive being in the center of a fast reactor

develop on-line fission product removal.

All these things are not included in Helion fusion engine development program and in fact will require more development work than LFTR. LFTR has the advantage of being similar to MSRE and hence has much development work already completed. I don't believe anyone has built a lead cooled, fluid fuel reactor yet at any size. 

One of the biggest challenges with LFTR is the lifetime of the first wall. In our case, that first wall is on the outer perimeter of the reactor chamber where it sees around 5% of the neutrons. In the reactor proposed in the paper above the first wall is between the fusion and fission reactors. It sees the full neutron flux from the fusion machine (and those are very fast so they more damaging to the wall). In addition it is near the heart of the fission reactor where it will see a large fission neutron flux. 

This is not to say he should stop work. Solving the energy problem is a very high value proposition worthy of several parallel efforts. But you will not have a power producing reactor for $40M using this approach. He hopes to build a fusion engine with break even power for this money scaled to supply 1/20th the neutrons used in a 1GWe reactor. 

You still have the expense of developing the fission reactor - that is not included in any of his costs. The fission reactor is the one that supplies 95% of the neutrons and all of the output power available to sell.

From Axil (how the first wall problem can be avoided with intermittent operation and easy and frequent swaps of the first wall):

If aneutronic B11-H fusion is not practical from either a technical or economic standpoint anytime in the near future, then D-T fusion is best served by a fusion/fission hybrid concept, and the Hilion reactor topology is well positioned for this approach.

Any big project should be developed in well thought out, mutually supportive and orchestrated phases. The thorium fusion hybrid should conform to this type of development strategy. The first phase should be the development of a U233 fuel factory. The first market would be existing Light Water Reactors and the new AHTR pebble reactors.

The price for the U233 would be well below the current U235 equivalent price. I think that such a fusion/fission fuel factory is very price competitive and is capable of producing U233 very well below this current $70 lbs yellowcake equivalent. The price of yellow cake has varied from $15 to $137 per pound recently and currently it is about $70 per pound.

32 page pdf that discusses makig lower priced uranium fuel

The best type of fusion/fission hybrid has a very small zone of fusion preferably a point source. The Hilion reactor has this very important feature and because of the small size of the fusion zone it facilitates an all inclosing blanket with almost perfect closure. Because of the ideal efficiency of its almost perfect liquid blanket envelope, I can see this type of subcritical reactor producing about 5500 kgs of pure U232/U233 per year. Very few neutrons would be wasted. Beryllium in the blanket would almost double the production of the fusion neutron flux. To maximize U233 production, no lithium should be included in the blanket. Tritium would come from the waste flow of its dependent parasitic fission reactors; its customers.

The reactor does not need heat exchangers of turboelectric generators; it can dump the heat produced by fusion (typically 10 megawatts) to the air so a thermal power circuit wound not need to be developed or deployed. Because it is subcritical, it would not need a containment structure either. 

If the protactinium is removed from the liquid fluoride beryllium/thorium blanket through on-line blanket salt reprocessing immediately after its creation, no fission heat would be produced by U233 fission. 

Since this hybrid does not need to produce electric power or connect to the grid, this hybrid can operate intermittingly to allow frequent change out of its first wall. Such a diamond pipe can be replaced in a matter of hours. A coating of lithium hydride on the inside of this first wall diamond pipe might greatly reduce alpha particle damage. 

I believe that this is the development strategy currently envisioned for the Helion fusion engine development program. 

If the U233 can be produced with a 1% or greater U232 content, then no U238 denaturing would be required by IAEA rules. This highly enriched and proliferation proof U232/U233 nuclear fuel would make light water reactors and AHTR very clean and eliminate the waste problem associated with the uranium fuel cycle. This alone would be a big selling point for the thorium/fusion hybrid and get the thorium fuel cycle off at a run.

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;nuclear,fission,thorium,fusion,energy,helion energy;Development Path for Helion Energy for 2010 and 2012;3089110 • Related • •

* Google Using Dwave Systems Quantum Computer as a binary classifier of images - http://feedblitz.com/r.asp?l=42939635&f=64651&u=2987253&c=0

Google is researching Quantum Computer Algorithms using Dwave Systems Quantum Computers for a binary classifier of images

At the Neural Information Processing Systems conference (NIPS 2009), we show the progress we have made. We demonstrate a detector that has learned to spot cars by looking at example pictures. It was trained with adiabatic quantum optimization using a D-Wave C4 Chimera chip. There are still many open questions but in our experiments we observed that this detector performs better than those we had trained using classical solvers running on the computers we have in our data centers today. Besides progress in engineering synthetic intelligence we hope that improved mastery of quantum computing will also increase our appreciation for the structure of reality as described by the laws of quantum physics.

A new type of machine, a so-called quantum computer, can help here. Quantum computers take advantage of the laws of quantum physics to provide new computational capabilities. While quantum mechanics has been foundational to the theories of physics for about a hundred years the picture of reality it paints remains enigmatic. This is largely because at the scale of our every day experience quantum effects are vanishingly small and can usually not be observed directly. Consequently, quantum computers astonish us with their abilities. Let’s take unstructured search as an example. Assume I hide a ball in a cabinet with a million drawers. How many drawers do you have to open to find the ball? Sometimes you may get lucky and find the ball in the first few drawers but at other times you have to inspect almost all of them. So on average it will take you 500,000 peeks to find the ball. Now a quantum computer can perform such a search looking only into 1000 drawers. This mind boggling feat is known as Grover’s algorithm.

Over the past three years a team at Google has studied how problems such as recognizing an object in an image or learning to make an optimal decision based on example data can be made amenable to solution by quantum algorithms. The algorithms we employ are the quantum adiabatic algorithms discovered by Edward Farhi and collaborators at MIT. These algorithms promise to find higher quality solutions for optimization problems than obtainable with classical solvers.

Training a Large Scale Classifier with the Quantum Adiabatic Algorithm (14 page pdf)

In a previous publication we proposed discrete global optimization as a method to train a strong binary classifier constructed as a thresholded sum over weak classifiers. Our motivation was to cast the training of a classifier into a format amenable to solution by the quantum adiabatic algorithm. Applying adiabatic quantum computing (AQC) promises to yield solutions that are superior to those which can be achieved with classical heuristic solvers. Interestingly we found that by using heuristic solvers to obtain approximate solutions we could already gain an advantage over the standard method AdaBoost. In this communication we generalize the baseline method to large scale classifier training. By large scale we mean that either the cardinality of the dictionary of candidate weak classifiers or the number of weak learners used in the strong classifier exceed the number of variables that can be handled effectively in a single global optimization. For such situations we propose an iterative and piecewise approach in which a subset of weak classifiers is selected in each iteration via global optimization. The strong classifier is then constructed by concatenating the subsets of weak classifiers. We show in numerical studies that the generalized method again successfully competes with AdaBoost. We also provide theoretical arguments as to why the proposed optimization method, which does not only minimize the empirical loss but also adds L0-norm regularization, is superior to versions of boosting that only minimize the empirical loss. By conducting a Quantum Monte Carlo simulation we gather evidence that the quantum adiabatic algorithm is able to handle a generic training problem efficiently.

NIPS 2009 Demonstration: Binary Classification using Hardware Implementation of Quantum Annealing (19 page pdf)

Previous work [NDRM08, NDRM09] has sought the development of binary classifiers that exploit the ability to better solve certain discrete optimization problems with quantum annealing. The resultant training algorithm was shown to offer benefits over competing binary classifiers even when the discrete optimization problems were solved with software heuristics. In this progress update we provide first results on training using a physical implementation of quantum annealing for black-box optimization of Ising objectives. We successfully build a classifier for the detection of cars in digital images using quantum annealing in hardware. We describe the learning algorithm and motivate the particular regularization we employ. We provide results on the efficacy of hardware-realized quantum annealing, and compare the final classifier to software trained variants, and a highly tuned version of AdaBoost

We test QBoost to develop a detector for cars in digital images. The training and test sets consist of 20 000 images with roughly half the images in each set containing cars and the other half containing city streets and buildings without cars. The images containing cars are human-labeled ground truth data with tight bounding boxes drawn around each car and grouped by positions (e.g. front, back, side, etc.) For demonstration purposes, we trained a single detector channel only using the side-view ground truth images. The actual data seen by the training system is obtained by randomly sampling subregions of the input training and test data to obtain 100 000 patches for each data set. Before presenting results on the performance of the strong classifier we characterize the optimization performance of the hardware.

Previous experiments on other data sets using software heuristics for QUBO solving have shown that performance beyond AdaBoost can typically be obtained. Presumably the improvements are due to the explicit regularization that QBoost employs. We would hope that QBoost can be made to outperform even the highly tuned boosting benchmark we have employed here.

We mention that the experiments presented here were not designed to test the quantumness of the hardware. Results of such tests will be reported elsewhere.

 • Email to a friend • Article Search - http://www.feedblitz.com/f/f.fbz?Search=64651;175594;quantum annealing,quantum computer,future,qubits,google,adiabatic quantum computer;Google Using Dwave Systems Quantum Computer as a binary classifier of images;3089110 • Related • •

* More Recent Articles

- Iraq Oil Capacity could Reach 12 million barrels per day in 6 years
- Magneto Inertial Fusion
- Next generation of retinal implants Has 17 Times Higher Resolution
- Improved Cooling for 3D Microchips With Signicant 3D Chip Deployment Expected 2015-2020
- Argonne Labs Working to Control Casimir Force
________________________________________________________
Unsubscribe from all current and future newsletters powered by FeedBlitz - http://www.feedblitz.com/f/f.fbz?EmailRemove=_Mjk4NzI1M3x8ZXVnZW5AbGVpdGwub3JnfDMwODkxMTA=_
----- End forwarded message -----
-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE