The new method creates truly random numbers with less computational effort than other methods, which could facilitate significantly higher levels of security for everything from consumer credit card transactions to military communications.

Computer science professor David Zuckerman and graduate student Eshan Chattopadhyay will present research about their method in June at the annual Symposium on Theory of Computing (STOC), the Association for Computing Machinery's premier theoretical computer science conference. An invitation to present at the conference is based on a rigorous peer review process to evaluate the work's correctness and significance. Their paper will be one of three receiving the STOC Best Paper Award.

Chattopadhyay and Zuckerman publicly released a draft paper describing their method for making random numbers in an online forum last year (http://eccc.hpi-web.de/report/2015/119/). In a field more accustomed to small, incremental improvements, the computer science community hailed the method, suggesting that, compared with earlier methods, this one is light years ahead. Oded Goldreich, a professor of computer science at the Weizmann Institute of Science in Israel, commented that even if it had only been a moderate improvement over existing methods, it would have justified a "night-long party."

The new method takes two weakly random sequences of numbers and turns them into one sequence of truly random numbers. Weakly random sequences, such as air temperatures and stock market prices sampled over time, harbor predictable patterns. Truly random sequences have nothing predictable about them, like a coin toss.

The new research seems to defy that old adage in computer programming, "Garbage in, garbage out." In fact, it's the latest, most powerful addition to a class of methods that Zuckerman pioneered in the 1990s called randomness extractors.

Previous versions of randomness extractors were less practical because they either required that one of the two source sequences be truly random (which presents a chicken or the egg problem) or that both source sequences be close to truly random. This new method sidesteps both of those restrictions and allows the use of two sequences that are only weakly random.

An important application for random numbers is in generating keys for data encryption that are hard for hackers to crack. Data encryption is critical for making secure credit card purchases and bank transactions, keeping personal medical data private and shielding military communications from enemies, among many practical applications.

Zuckerman says that although there are already methods for producing high-quality random numbers, they are very computationally demanding. His method produces higher quality randomness with less effort.

Their paper shows how to generate only one truly random number -- akin to one coin toss -- but Zuckerman's former student Xin Li has already demonstrated how to expand it to create sequences of many more random numbers.

The website where Zuckerman and Chattopadhyay posted their draft last summer, called the Electronic Colloquium on Computational Complexity, allows researchers to share their work and receive feedback before publishing final versions in journals or at conferences. Computer scientists and mathematicians have been carefully reviewing the article, providing suggestions and even extending the method to make it more powerful.

They found that salts used in the liquid in the batteries make a big difference. When a salt called LiTFSI is packed in the liquid, a test battery can hold most of its charge for more than 200 uses. The LiTFSI helps bind up lithium atoms and sulfur on the electrode but quickly releases them. In contrast, a similar liquid ties up the lithium and sulfur but doesn't release it. The result is an electrode that quickly degrades; the battery fades after a few dozen uses.

One of the concerns with electric cars is long, lonely stretches of highway. Drivers don't want to be stranded between charging stations, and this concern can factor into their decision to buy lower emission vehicles. The results of this study add another important page into the design guide for high-energy lithium-sulfur batteries.

To determine the influence of electrolytes in lithium-sulfur batteries, the team did experiments with both LiTFSI and a similar electrolyte, called LiFSI, which has less carbon and fluoride. After continually measuring the amount of energy that the battery held and released, the team did a post-mortem analysis to study the electrodes. They did this work using instruments at DOE's EMSL, an Office of Science scientific user facility.

They discovered that with the LiTFSI, the electrode's lithium atoms became bound up with sulfur. The result is lithium sulfide (LiSx) forming on the electrode's surface. With LiFSI, lithium sulfate (LiSOx) formed. By calculating the strength with which the compounds clung to the lithium, they found that the lithium sulfide easily broke apart to release the lithium. However, the lithium sulfate was hard to separate. The oxygen in the lithium sulfate was the culprit.

"By conducting a macroscopic compositional analysis combined with simulations, we can see which bonds are easily broken and what will happen from there," said Dr. Ji-Guang (Jason) Zhang, who led the study at the national laboratory. "This process lets us identify the electrolytes behavior, guides us to design a better electrolyte, and improve the cycle life of lithium-sulfur batteries."

For the researchers, the next step is developing an electrolyte additive that forms a protective layer on the lithium anode's surface, protecting it from the electrolyte.

Tufts University biologists using a frog model have demonstrated for the first time that it is possible to prevent tumors from forming and normalize tumors after they have formed by using light to control electrical signaling among cells. The work, which appears online in Oncotarget on March 16, 2016 is the first reported use of optogenetics to specifically manipulate bioelectrical signals to both prevent and cause regression of tumors induced by oncogenes.

Frogs are a good model for basic science research into cancer because tumors in frogs and mammals share many of the same characteristics. These include rapid cell division, tissue disorganization, increased vascular growth, invasiveness and cells that have an abnormally positive internal electric voltage.

 

Virtually all healthy cells maintain a more negative voltage in the cell interior compared with the cell exterior; the opening and closing of ion channels in the cell membrane can cause the voltage to become more positive (depolarizing the cell) or more negative (polarizing the cell). Tumors can be detected by their abnormal bioelectrical signature before they are otherwise apparent.

"These electrical properties are not merely byproducts of oncogenic processes. They actively regulate the deviations of cells from their normal anatomical roles towards tumor growth and metastatic spread," said senior and corresponding author Michael Levin, Ph.D., who holds the Vannevar Bush chair in biology and directs the Center for Regenerative and Developmental Biology at Tufts School of Arts and Sciences. "Discovering new ways to specifically control this bioelectrical signaling could be an important path towards new biomedical approaches to cancer."

Lead author Brook Chernet, Ph.D., former post-doctoral associate in the Levin laboratory, injected cells in Xenopus laevis embryos with RNA encoding a mutant RAS oncogene known to cause cancer-like growths. The researchers also expressed and activated either a blue light-activated, positively charged ion channel, ChR2D156A, or a green light-activated proton pump, Archaerhodopsin (Arch), both of which hyperpolarize frog embryonic cells, thereby inducing an electric current that caused the cells to go from a cancer-like depolarized state to a normal, more negative polarized state. Activation of both agents significantly lowered the incidence of tumor formation and also increased the frequency with which tumors regressed into normal tissue.

The use of light to control ion channels has been a ground-breaking tool in research on the nervous system and brain, but optogenetics had not yet been applied to cancer.

"This provides proof of principle for a novel class of therapies which use light to override the action of oncogenic mutations," said Levin. "Using light to specifically target tumors would avoid subjecting the whole body to toxic chemotherapy or similar reagents."

Machine learning systems are everywhere. They predict the weather, forecast earthquakes, provide recommendations based on the books and movies we like, and even apply the brakes on our cars when we're not paying attention.

To do this, software programs in these systems calculate predictive relationships from massive amounts of data. The systems identify these predictive relationships using advanced algorithms -- a set of rules for solving math problems -- and "training data." This data is then used to construct the models and features that enable a system to determine the latest best-seller you wish to read or to predict the likelihood of rain next week.

This intricate process means that a piece of raw data often goes through a series of computations in a system. The computations and information derived by the system from that data together form a complex propagation network called the data's "lineage." The term was coined by Yinzhi Cao, an assistant professor of computer science and engineering, and his colleague, Junfeng Yang of Columbia University, who are pioneering a novel approach to make learning systems forget.

Considering how important this concept is to increasing security and protecting privacy, Cao and Yang believe that easy adoption of forgetting systems will be increasingly in demand. The two researchers have developed a way to do it faster and more effectively than can be done using current methods.

Their concept, called "machine unlearning," is so promising that Cao and Yang have been awarded a four-year, $1.2 million National Science Foundation grant to develop the approach.

"Effective forgetting systems must be able to let users specify the data to forget with different levels of granularity," said Cao, a principal investigator on the project. "These systems must remove the data and undo its effects so that all future operations run as if the data never existed."

Increasing security and privacy protection
There are a number of reasons why an individual user or service provider might want a system to forget data and its complete lineage. Privacy is one.

After Facebook changed its privacy policy, many users deleted their accounts and the associated data. The iCloud photo hacking incident in 2014 -- in which hundreds of celebrities' private photos were accessed via Apple's cloud services suite -- led to online articles teaching users how to completely delete iOS photos including the backups. New research has revealed that machine learning models for personalized medicine dosing leak patients' genetic markers. Only a small set of statistics on genetics and diseases are enough for hackers to identify specific individuals, despite cloaking mechanisms.

Naturally, users unhappy with these newfound risks want their data, and its influence on the models and statistics, to be completely forgotten.

Security is another reason. Consider anomaly-based intrusion detection systems used to detect malicious software. In order to positively identify an attack, the system must be taught to recognize normal system activity. Therefore the security of these systems hinges on the model of normal behaviors extracted from the training data. By polluting the training data, attackers pollute the model and compromise security. Once the polluted data is identified, the system must completely forget the data and its lineage in order to regain security.

Widely used learning systems such as Google Search are, for the most part, only able to forget a user's raw data -- and not the data's lineage -- upon request. This is problematic for users who wish to ensure that any trace of unwanted data is removed completely, and it is also a challenge for service providers who have strong incentives to fulfill data removal requests and retain customer trust.

Service providers will increasingly need to be able to remove data and its lineage completely to comply with laws governing user data privacy, such as the "right to be forgotten" ruling issued in 2014 by the European Union's top court. In October 2014, Google removed more than 170,000 links to comply with the ruling, which affirmed users' right to control what appears when their names are searched. In July 2015, Google said it had received more than a quarter-million such requests.

Breaking down dependencies

Building on work that was presented at a 2015 IEEE Symposium and then published, Cao and Yang's "machine unlearning" method is based on the fact that most learning systems can be converted into a form that can be updated incrementally without costly retraining from scratch.

Their approach introduces a layer of a small number of summations between the learning algorithm and the training data to eliminate dependency on each other. So, the learning algorithms depend only on the summations and not on individual data. Using this method, unlearning a piece of data and its lineage no longer requires rebuilding the models and features that predict relationships between pieces of data. Simply recomputing a small number of summations would remove the data and its lineage completely -- and much faster than through retraining the system from scratch.

Cao believes he and Yang are the first to establish the connection between unlearning and the summation form.

And, it works. Cao and Yang tested their unlearning approach on four diverse, real-world systems: LensKit, an open-source recommendation system; Zozzle, a closed-source JavaScript malware detector; an open-source OSN spam filter; and PJScan, an open-source PDF malware detector.

The success of these initial evaluations has set the stage for the next phases of the project, which include adapting the technique to other systems and creating verifiable machine unlearning to statistically test whether unlearning has indeed repaired a system or completely wiped out unwanted data.

In their paper's introduction, Cao and Yang say that "machine unlearning" could play a key role in enhancing security and privacy and in our economic future:

"We foresee easy adoption of forgetting systems because they benefit both users and service providers. With the flexibility to request that systems forget data, users have more control over their data, so they are more willing to share data with the systems. More data also benefit the service providers, because they have more profit opportunities and fewer legal risks.

"We envision forgetting systems playing a crucial role in emerging data markets where users trade data for money, services, or other data because the mechanism of forgetting enables a user to cleanly cancel a data transaction or rent out the use rights of her data without giving up the ownership." 

Deepika Raj and Jung E Ha-Brookshire of the University of Missouri interviewed people employed in the WT industry -- with its unique collaboration between apparel and technology specialists -- to find out how creating the devices changed the way they worked. They also wanted to find out what the designers themselves thought of the items they were creating.

As Raj and Ha-Brookshire explain: "Despite the significance and popularity of WT in today's market environment, little research has been done to understand what WT employees think about WT when developing new products. We also know little about the knowledge-creation processes that WT employees may go through to develop innovative WT products and specific work environments that are ideal for WT innovation.

Perhaps unsurprisingly, the respondents were very positive about the benefits of WT. Replies like 'it makes you do more with less', 'creates fun and meaningful solutions to everyday problems' and 'solves multiple problems' were common.

The notion of WT giving people superpowers also cropped up repeatedly. "For the less-able bodied, [one of the respondents] believed WT could help them hear, see, or speak better. For able-bodied people … WT could help them do things they could not do before," the authors note. "In both cases, WT gave a sort of superpower to human beings."

The workers were also clear on what made good WT: it had to be easy to use, 'fit the curve of the body', improve connectivity with the outside world and, ideally, solve problems for the user.

Perhaps fittingly, working on WT also changed the way the WT specialists themselves worked. As good WT needs to draw on different knowledge bases, the workers found themselves collaborating more and understanding better both the terms and working practices of other disciplines. "New knowledge was created within WT firms through the process of socialisation by personal communications, internalisation and by trial and error," the authors observe. And just as WT can empower its users, 'innovation in WT firms seems to be possible by empowering the work force and giving them a productive atmosphere to foster creativity and collaborations.'

Certainly for the WT workers themselves, it would appear that WT can change lives in more ways than one.

A recent observational campaign involving more than two dozen optical telescopes and NASA's space based SWIFT X-ray telescope allowed a team of astronomers to measure very accurately the rotational rate of one of the most massive black holes in the universe. The rotational rate of this massive black hole is one third of the maximum spin rate allowed in General Relativity. This 18 billion solar mass heavy black hole powers a quasar called OJ287 which lies about 3.5 billion light years away from Earth. Quasi-stellar radio sources or `quasars' for short, are the very bright centers of distant galaxies which emit huge amounts of electro-mag

netic radiation due to the infall of matter into their massive black holes.

This quasar lies very close to the apparent path of the Sun's motion on the celestial sphere as seen from Earth, where most searches for asteroids and comets are conducted. Therefore, its optical photometric measurements already cover more than 100 years. A careful analysis of these observations show that OJ 287 has produced quasi-periodic optical outbursts at intervals of approximately 12 years dating back to around 1891. Additionally, a close inspection of newer data sets reveals the presence of double-peaks in these outbursts.

These deductions prompted Prof. Mauri Valtonen of University of Turku, Finland and his collaborators to develop a model that requires the quasar OJ287 to harbour two unequal mass black holes. Their model involves a massive black hole with an accretion disk (a disk of interstellar material formed by matter falling into objects like black holes) while the comparatively smaller black hole revolves around it. The quasar OJ287 is visible due to the slow accretion of matter, present in the accretion disk, onto the largest black hole. Additionally, the small black hole passes through the accretion disk during its orbit which causes the disk material to heat up to very high temperatures. This heated material flows out from both sides of the accretion disk and radiates strongly for weeks. This causes peaks in the brightness, and the double peaks arise due to the ellipticity of the orbit, as shown in the figure.

The binary black hole model for OJ287 implies that the smaller black hole's orbit should rotate, and this changes where and when the smaller hole impacts the accretion disk. This effect arises from Einstein's General Theory of Relativity and its precessional rate depends mainly on the two black hole masses and the rotation rate of the more massive black hole. In 2010, Valtonen and collaborators used eight well timed bright outbursts of OJ287 to accurately measure the precession rate of the smaller hole's orbit. This analysis revealed for the first time the rotation rate of the massive black hole along with accurate estimates for the masses of the two black holes. This was possible since the smaller black hole's orbit precess at an incredible 39 degrees per individual orbit. The General Relativistic model for OJ287 also predicted that the next outburst could occur around the time of GR Centenary, 25 November 2015, which marks the 100th anniversary of Einstein's General Theory of Relativity.

An observational campaign was therefore launched to catch this predicted outburst. The predicted optical flare began around November 18, 2015 and reached its maximum brightness on December 4, 2015. It is the timing of this bright outburst that allowed Valtonen and his co-workers to directly measure the rotation rate of the more massive black hole to be one third of the maximum spin rate allowed in General Relativity. In other words, its Kerr parameter is accurately measured to be 0.31 and its maximum allowed value in General Relativity is one. In comparison, the Kerr parameter of the final black hole associated with the first ever direct detection of gravitational waves is only estimated to be below 0.7.

The observations leading to accurate spin measurement have been made due to the collaboration of a number of optical telescopes in Japan, South Korea, India, Turkey, Greece, Finland, Poland, Germany, UK, Spain, USA and Mexico. The effort, led by Staszek Zola of Poland, involved close to 100 astronomers from these countries. Interestingly, a number of key participants were amateur astronomers who operate their own telescopes. Valtonen's team that developed and contributed to the spinning binary black hole model include theoretical astrophysicist A. Gopakumar from TIFR, India, and Italian X-Ray astronomer Stefano Ciprini who obtained and analyzed the X-ray data.

The occurrence of the predicted optical outburst of OJ287 also allowed the team to confirm the loss of orbital energy to gravitational waves within two percent of General Relativity's prediction. This provides the first indirect evidence for the existence of a massive spinning black hole binary emitting gravitational waves. This is encouraging news for the Pulsar Timing Array efforts that will directly detect gravitational waves from such systems in the near future. Therefore, the present optical outburst of OJ287 makes a fitting contribution to the centenary celebrations of General Relativity and adds to the excitement of the first direct observation of a transient gravitational wave signal by LIGO.

The foldable object that was created by researchers

Imagine a house that could fit in a backpack or a wall that could become a window with the flick of a switch.

Harvard researchers have designed a new type of foldable material that is versatile, tunable and self actuated. It can change size, volume and shape; it can fold flat to withstand the weight of an elephant without breaking, and pop right back up to prepare for the next task.


The research was lead by Katia Bertoldi, the John L. Loeb Associate Professor of the Natural Sciences at the John A. Paulson School of Engineering and Applied Sciences (SEAS), James Weaver, Senior Research Scientist at the Wyss Institute for Biologically Inspired Engineering at Harvard University and Chuck Hoberman, of the Graduate School of Design. It is described in Nature Communications.


"We've designed a three-dimensional, thin-walled structure that can be used to make foldable and reprogrammable objects of arbitrary architecture, whose shape, volume and stiffness can be dramatically altered and continuously tuned and controlled," said Johannes T. B. Overvelde, graduate student in Bertoldi's lab and first author of the paper.


The structure is inspired by an origami technique called snapology, and is made from extruded cubes with 24 faces and 36 edges. Like origami, the cube can be folded along its edges to change shape. The team demonstrated, both theoretically and experimentally, that the cube can be deformed into many different shapes by folding certain edges, which act like hinges. The team embedded pneumatic actuators into the structure, which can be programmed to deform specific hinges, changing the cube's shape and size, and removing the need for external input.


The team connected 64 of these individual cells to create a 4x4x4 cube that can grow, and shrink, change its shape globally, change the orientation of its microstructure and fold completely flat. As the structure changes shape, it also changes stiffness -- meaning one could make a material that's very pliable or very stiff using the same design. These actuated changes in material properties adds a fourth dimension to the material.


"We not only understand how the material deforms, but also have an actuation approach that harnesses this understanding," said Bertoldi. "We know exactly what we need to actuate in order to get the shape we want."


The material can be embedded with any kind of actuator, including thermal, dielectric or even water.


"The opportunities to move all of the control systems onboard combined with new actuation systems already being developed for similar origami-like structures really opens up the design space for these easily deployable transformable structures," said Weaver.


"This structural system has fascinating implications for dynamic architecture including portable shelters, adaptive building facades and retractable roofs," said Hoberman. "Whereas current approaches to these applications rely on standard mechanics, this technology offers unique advantages such as how it integrates surface and structure, its inherent simplicity of manufacture, and its ability to fold flat."


"This research demonstrates a new class of foldable materials that is also completely scalable," Overvelde said, " It works from the nanoscale to the meter-scale and could be used to make anything from surgical stents to portable pop-up domes for disaster relief."

The "birth" of neurons has been witnessed in a live brain for the first time.

Neurons provide the crucial mechanism that allows the brain to operate; 'born' in the hippocampus, they are responsible for processing memories and keeping related, but separate, memories distinct from one another.

 What wasn't always clear is how -- or even if -- new neurons can be made in the brain. Though it now seems clear that they are, that process had never been witnessed first hand -- the brain is complex and delicate, and neurons are extremely small and intricate.

To get around those restrictions, a team at Columbia University Medical Center in New York used several techniques to observe the birth of the neurons -- first implanting a miniature microscope into the brains of mice, and then chaning the mice so that their new neurons would glow enough to be picked up by that device. Once prepared, the mice were placed on a treadmill and ran while they experienced slightly different sights and smells, and were given electric shocks to create an association between the stimuli and the unpleasant experience.


When the neurons were then 'turned off', the mice were no longer able to tell between cues that were neutral and cues that were unpleasant. This, the team says, suggests that the new neurons were responsible for telling the difference between similar memories.

This result was replicated when the mice were placed into identical chambers, in which they had received shocks and rewards at different times.

Because the skill responsible for separating memories is reduced in people with anxiety disorders, the team hope that their research will prove useful for further investigations of depression and anxiety. Many antidepressants "stimulate the production of new hippocampus neurons", the team says, so their new discovery may provide new insight into the link between the neurons and mental illness.

By using WiFi data collected on campus, EPFL researchers were able to analyze students' motivations in a fundamental activity: eating. More broadly, this method offers a low-cost way of studying pedestrians' comings, goings and destinations.

On a campus with a population of 13,000, there is no shortage of eateries. How do people choose one? Setting aside the traditional survey approach, EPFL researchers looked at the traces left by pedestrians as they passed WiFi access points. These virtual stones deposited by streams of people are worth their weight in gold when it comes to understanding and predicting the behavior of people who use the pedestrian infrastructure. Provided we can read the data, that is.


Antonin Danalet, a researcher in the Transport and Mobility Laboratory, chose EPFL's campus as his testing ground. He recorded more than two million points picked up by the campus's 789 WiFi antennas. Data was collected over 10 days in mid-2012. "WiFi traces are cheap and easy to obtain, but they do not mean much on their own," said the researcher. For one, they are vague. While they indicate how much distance was covered in how much time, they do not on first glance give any insight into the route people took or what they did once at their destination. The aim of Danalet's thesis was to make this data talk.

With the help of the school's IT service, he was able to link the points he collected with individuals – around 2,000 in all – in order to give them meaning. "The data was anonymized," he said. "I only knew if the traces came from an EPFL student or staff member. If it was a student, I also knew the section and year."

To understand the individuals' successive destinations, Danalet merged WiFi location data with map data. By using the campus map and its pedestrian walkways, he was able to reconstruct routes and identify when the pedestrians reached their destination, how long they spent there and what they did. The researcher also superimposed data on the use of the campus's rooms – by taking the occupancy rate of classrooms as determined by class schedules, for example – as well as data on the number of purchases made at the various eateries, which gives the number of people who pulled out their wallet at a given time on a given day.

 

"We have statistics and numbers on people who drive and take the train, but pedestrian behavior is often a mystery," said the researcher. "These observations would come in very handy for understanding the use of pedestrian infrastructure at music festivals, museums and hospitals, for example."
 

To verify the accuracy of his model, Danalet compared it to the comings and goings of an actual person: himself. "This allowed me to correct any biases in the model," said Danalet. Another cross-check: "We compared the WiFi traces to the results of a survey asking students when they arrive on campus and when they leave." The result? According to the WiFi traces, students arrive in the morning on campus around half an hour later than they report in the survey. "Do they turn on their phone a half hour after arriving on campus, or were they trying to look good for the survey?" wonders the researcher. "We have no idea. But psychologists refer to the second hypothesis as social desirability bias."

In addition to people's comings and goings, the WiFi traces can also provide some insight into the choices they make. For example, why they choose one restaurant over another on a campus with 21 choices (back in 2012). The researcher assigned a score to each eatery on campus. He included various criteria, such as price, size, proximity, the annual quality rating, the food served and the opening hours. Danalet then applied the WiFi data – such as where the architecture student prefers to nosh – to identify the most important factors. Not surprisingly on a campus, price is a key factor. But it's not the only one. Proximity is also very important, as is size: the larger the eatery, the faster the line. The type of food does not have much bearing, it turns out.

Danalet offers this conclusion: "This model is appealing because it can predict changes in what people do and where they go as a function of changes in the choices presented to them. What will happen if the price of the daily special goes up? Where will the cost-conscious go? And what sort of impact will a new eatery have? This type of pedestrian behavior model is important for designing – or improving – pedestrian infrastructure at places like train stations, airports and shopping centers. These models can also help identify the best spot for things such as a cash machine or bakery."