When machine learning packs an economic punch

A new study co-authored by an MIT economist shows that improved translation software can significantly boost international trade online — a notable case of machine learning having a clear impact on economic activity.

The research finds that after eBay improved its automatic translation program in 2014, commerce shot up by 10.9 percent among pairs of countries where people could use the new system.   

“That’s a striking number. To have it be so clear in such a short amount of time really says a lot about the power of this technology,” says Erik Brynjolfsson, an MIT economist and co-author of a new paper detailing the results.

To put the results in perspective, he adds, consider that physical distance is, by itself, also a significant barrier to global commerce. The 10.9 percent change generated by eBay’s new translation software increases trade by the same amount as “making the world 26 percent smaller, in terms of its impact on the goods that we studied,” he says.

The paper, “Does Machine Translation Affect International Trade? Evidence from a Large Digital Platform,” appears in the December issue of Management Science. The authors are Brynjolfsson, who is the Schussel Family Professor of Management Science at the MIT Sloan School of Management, and Xiang Hui and Meng Liu, who are both assistant professors in the Olin Business School at Washington University in St. Louis.

Just cause

To conduct the study, the scholars examined what happened after eBay, in 2014, introduced its new eBay Machine Translation (eMT) system — a proprietary machine-learning program that, by several objective measures, significantly improved translation quality on eBay’s site. The new system initially was focused on English-Spanish translations, to facilitate trade between the United States and Latin America

Previously, eBay had used Bing Translator to render the titles of objects for sale. By one evaluation measure, called the Human Acceptance Rate (HAR), in which three experts accept or reject translations, the eMT system increased the number of acceptable Spanish-language item titles on eBay from 82 percent to 90 percent.

Using administrative data from eBay, the researchers then examined the volume of trade on the platform, within countries, after the eMT system went into use. Other factors being equal, the study showed that the new translation system not only had an effect on sales, but that trade increased by 1.06 percent for each additional word in the titles of items on eBay.

That is a substantial change for a commerce platform on which, as the paper notes, items for sale often have long, descriptive titles such as “Diamond-Cut Stackable Thin Wedding Ring New .925 Sterling Silver Band Sizes 4-12,” or “Alpine Swiss Keira Women’s Trench Coast Double Breasted Wool Jacket Belted.” In those cases, making the translation clearer helps potential buyers understand exactly what they might be purchasing.

Given the study’s level of specificity, Brynjolfsson calls it “a really fortunate natural experiment, with a before-and-after that sharply distinguished what happened when you had machine translation and when you didn’t.”

The structure of the study, he adds, has enabled the researchers to say with confidence that the new eBay program, and not outside factors, directly generated the change in trade volume among affected countries.

“In economics, it’s often hard to do causal analyses and prove that A caused B, not just that A was associated with B,” says Brynjolfsson. “But in this case, I feel very comfortable using causal language and saying that improvement in machine translation caused the increase in international trade.”

Larger puzzle: The productivity issue

The genesis of the paper stems from an ongoing question about new technology and economic productivity. While many forms of artificial intelligence have been developed and expanded in the last couple of decades, the impact of AI, including things like machine-translation systems, has not been obvious in economics statistics.

“There’s definitely some amazing progress in the core technologies, including in things like natural language processing and translation,” Brynjolfsson says. “But what’s been lacking has been evidence of an economic impact, or business impact. So that’s a bit of a puzzle.”

When looking to see if an economic impact for various forms of AI could be measured, Brynjolfsson, Hui, and Liu thought machine translation “made sense, because it’s a relatively straightforward implementation,” Brynjolfsson adds. That is, better translations could influence economic activity, at least on eBay, without any other changes in technology occurring.

In this vein, the findings fit with a larger postulation Brynjolfsson has developed in recent years — that the adoption of AI technologies produces a “J-curve” in productivity. As Brynjolfsson has previously written, broad-ranging AI technologies nonetheless “require significant complementary investments, including business process redesign, co-invention of new products and business models, and investments in human capital” to have a large economic impact.

As a result, when AI technologies are introduced, productivity may appear to slow down, and when the complementary technologies are developed, productivity may appear to take off — in the “J-curve” shape.

So while Brynjolfsson believes the results of this study are clear, he warns against generalizing too much on the basis of this finding about the impact of machine learning and other forms of AI on economic activity. Every case is different, and AI will not always produce such notable changes by itself.

“This was a case where not a lot of other changes had to happen in order for the technology to benefit the company,” Brynjolfsson says. “But in many other cases, much more complicated, complementary changes are needed. That’s why, in most cases with machine learning, it takes longer for the benefits to be delivered.”

de MIT News https://ift.tt/38YFdFP

Exploring hip hop history with art and technology

A new museum is coming to New York City in 2023, the year of hip-hop’s 50th birthday, and an MIT team has helped to pave the way for the city to celebrate the legacy of this important musical genre — by designing unique creative experiences at the intersection of art, learning, and contemporary technology.

With “The [R]evolution of Hip Hop Breakbeat Narratives,” a team led by D. Fox Harrell, professor of digital media and artificial intelligence and director of the MIT Center for Advanced Virtuality, has created an art installation that takes museum-goers on an interactive, personalized journey through hip hop history.

The installation served as the centerpiece of an event held this month by leaders of the highly anticipated Universal Hip Hop Museum (UHHM), which will officially open in just a few years in the Bronx — the future home of the UHHM, and where many agree that the genre of hip hop music originated.

“Hip hop is much more than a musical genre. It is a global phenomenon, with a rich history and massive social and cultural impact, with local roots in the Bronx,” Harrell says. “As an educational center, the Universal Hip Hop Museum will have the power to connect people to the surrounding community.”

Harrell’s immersive art installation takes museum-goers on a journey through hip hop culture and history, from the 1970s to the present. However, not everyone experiences the installation in the same way. Using a computational model of users’ preferences and artificial intelligence technologies to drive interaction, the team of artists and computer scientists from the Center for Advanced Virtuality has created layered, personalized virtual experiences.

When approaching the exhibit, museum-goers are greeted by “The Elementals,” or novel characters named after the five elements of hip hop (MC, DJ, Breakdance, Graffiti Art, and Knowledge) that guide users and ask key questions — “What is your favorite hip hop song?” or “Which from this pair of lyrics do you like the most?” Based on those answers, the Elementals take users through their own personalized narrative of hip hop history.

Harrell developed the Elementals with professors John Jennings of the University of California at Riverside and Stacey Robinson of the University of Illinois — artists collectively known as Black Kirby. This visual aesthetic ties the work into the rich, imaginative cultures and iconography of the African diaspora.

Through these conversations with the Elementals they encounter, people can explore broad social issues surrounding hip hop, such as gender, fashion, and location. At the end of their journey, they can take home a personalized playlist of songs. 

“We designed the Breakbeat Narratives installation by integrating Microsoft conversational AI technologies, which made our user modeling more personable, with a music visualization platform from the TunesMap Educational Foundation,” Harrell says.

The exploration of social issues is about as close to the heart of Harrell’s mission in the Center for Advanced Virtuality as one can get. In the center, Harrell designs virtual technologies to stimulate creative expression, cultural analysis, and positive social change.

“We wanted to tell stories that pushed beyond stereotypical representations, digging into the complexities of both empowering and problematic representations that often coexist,” he says. “This work fits into our endeavor called the Narrative, Orality, and Improvisation Research (NOIR) Initiative that uses AI technologies to forward the art forms of diverse global cultures.”

Through this art project enabled by contemporary technologies, Harrell hopes that he has helped museum leadership to achieve their goal of celebrating hip-hop’s heritage and legacy.

“Now, people internationally can have a stake in this great art.”

de MIT News https://ift.tt/36Uf8Wv

Researchers produce first laser ultrasound images of humans

For most people, getting an ultrasound is a relatively easy procedure: As a technician gently presses a probe against a patient’s skin, sound waves generated by the probe travel through the skin, bouncing off muscle, fat, and other soft tissues before reflecting back to the probe, which detects and translates the waves into an image of what lies beneath.

Conventional ultrasound doesn’t expose patients to harmful radiation as X-ray and CT scanners do, and it’s generally noninvasive. But it does require contact with a patient’s body, and as such, may be limiting in situations where clinicians might want to image patients who don’t tolerate the probe well, such as babies, burn victims, or other patients with sensitive skin. Furthermore, ultrasound probe contact induces significant image variability, which is a major challenge in modern ultrasound imaging.

Now, MIT engineers have come up with an alternative to conventional ultrasound that doesn’t require contact with the body to see inside a patient. The new laser ultrasound technique leverages an eye- and skin-safe laser system to remotely image the inside of a person. When trained on a patient’s skin, one laser remotely generates sound waves that bounce through the body. A second laser remotely detects the reflected waves, which researchers then translate into an image similar to conventional ultrasound.

In a paper published today by Nature in the journal Light: Science and Applications, the team reports generating the first laser ultrasound images in humans. The researchers scanned the forearms of several volunteers and observed common tissue features such as muscle, fat, and bone, down to about 6 centimeters below the skin. These images, comparable to conventional ultrasound, were produced using remote lasers focused on a volunteer from half a meter away.

“We’re at the beginning of what we could do with laser ultrasound,” says Brian W. Anthony, a principal research scientist in MIT’s Department of Mechanical Engineering and Institute for Medical Engineering and Science (IMES), a senior author on the paper. “Imagine we get to a point where we can do everything ultrasound can do now, but at a distance. This gives you a whole new way of seeing organs inside the body and determining properties of deep tissue, without making contact with the patient.”

Anthony’s co-authors on the paper are lead author and MIT postdoc Xiang (Shawn) Zhang, recent doctoral graduate Jonathan Fincke, along with Charles Wynn, Matthew Johnson, and Robert Haupt of MIT’s Lincoln Laboratory.

Yelling into a canyon — with a flashlight

In recent years, researchers have explored laser-based methods in ultrasound excitation in a field known as photoacoustics. Instead of directly sending sound waves into the body, the idea is to send in light, in the form of a pulsed laser tuned at a particular wavelength, that penetrates the skin and is absorbed by blood vessels.

The blood vessels rapidly expand and relax — instantly heated by a laser pulse then rapidly cooled by the body back to their original size — only to be struck again by another light pulse. The resulting mechanical vibrations generate sound waves that travel back up, where they can be detected by transducers placed on the skin and translated into a photoacoustic image.

While photoacoustics uses lasers to remotely probe internal structures, the technique still requires a detector in direct contact with the body in order to pick up the sound waves. What’s more, light can only travel a short distance into the skin before fading away. As a result, other researchers have used photoacoustics to image blood vessels just beneath the skin, but not much deeper.

Since sound waves travel further into the body than light, Zhang, Anthony, and their colleagues looked for a way to convert a laser beam’s light into sound waves at the surface of the skin, in order to image deeper in the body. 

Based on their research, the team selected 1,550-nanometer lasers, a wavelength which is highly absorbed by water (and is eye- and skin-safe with a large safety margin).  As skin is essentially composed of water, the team reasoned that it should efficiently absorb this light, and heat up and expand in response. As it oscillates back to its normal state, the skin itself should produce sound waves that propagate through the body.

The researchers tested this idea with a laser setup, using one pulsed laser set at 1,550 nanometers to generate sound waves, and a second continuous laser, tuned to the same wavelength, to remotely detect reflected sound waves.  This second laser is a sensitive motion detector that measures vibrations on the skin surface caused by the sound waves bouncing off muscle, fat, and other tissues. Skin surface motion, generated by the reflected sound waves, causes a change in the laser’s frequency, which can be measured. By mechanically scanning the lasers over the body, scientists can acquire data at different locations and generate an image of the region.

“It’s like we’re constantly yelling into the Grand Canyon while walking along the wall and listening at different locations,” Anthony says. “That then gives you enough data to figure out the geometry of all the things inside that the waves bounced against — and the yelling is done with a flashlight.”

In-home imaging

The researchers first used the new setup to image metal objects embedded in a gelatin mold roughly resembling skin’s water content. They imaged the same gelatin using a commercial ultrasound probe and found both images were encouragingly similar. They moved on to image excised animal tissue — in this case, pig skin — where they found laser ultrasound could distinguish subtler features, such as the boundary between muscle, fat, and bone.

Finally, the team carried out the first laser ultrasound experiments in humans, using a protocol that was approved by the MIT Committee on the Use of Humans as Experimental Subjects. After scanning the forearms of several healthy volunteers, the researchers produced the first fully noncontact laser ultrasound images of a human. The fat, muscle, and tissue boundaries are clearly visible and comparable to images generated using commercial, contact-based ultrasound probes.

The researchers plan to improve their technique, and they are looking for ways to boost the system’s performance to resolve fine features in the tissue. They are also looking to hone the detection laser’s capabilities. Further down the road, they hope to miniaturize the laser setup, so that laser ultrasound might one day be deployed as a portable device.

“I can imagine a scenario where you’re able to do this in the home,” Anthony says. “When I get up in the morning, I can get an image of my thyroid or arteries, and can have in-home physiological imaging inside of my body. You could imagine deploying this in the ambient environment to get an understanding of your internal state.” 

This research was supported in part by the MIT Lincoln Laboratory Biomedical Line Program for the United States Air Force and by the U.S. Army Medical Research and Material Command's Military Operational Medicine Research Program.

de MIT News https://ift.tt/2rVRPwS

Israel Ruiz to step down as MIT’s executive vice president and treasurer

Israel Ruiz will step down next year as MIT’s executive vice president and treasurer, a position he has held since 2011. President L. Rafael Reif announced the news today in a letter to MIT faculty and staff.

“Widely respected across MIT, Israel is a brilliant strategic thinker whose commitment to excellence has advanced innovative solutions to complex challenges while earning broad support along the way,” Reif wrote. “His efforts have transformed many aspects of our campus to better serve and support the MIT community. Since my earliest days as provost, he has been among my most important advisors.”

Ruiz, who has played a key role over the past decade in numerous MIT initiatives to advance innovation and entrepreneurship — efforts ranging from MIT’s Kendall Square Initiative to the launches of MITx, edX, and The Engine — plans to devote the next chapter of his career to efforts outside of academia in which he can more directly drive innovation and impact.

“I am immensely proud of my work to help launch these transformative initiatives,” Ruiz says. “At the same time, considering the accomplishments of the last decade and my career at MIT, I’ve been contemplating a change over the last couple of years. I feel it is time for me to focus firsthand on opportunities that accelerate innovation in the way this community has inspired me to do.”

Ruiz expects to transition out of his role at MIT during the spring semester. In his letter to the community, President Reif indicated that he will work in the coming months with members of Ruiz’s senior team — including Vice President for Finance Glen Shor and Vice President for Campus Services and Stewardship Joe Higgins — to determine how best to allocate Ruiz’s responsibilities.

Ruiz arrived at MIT as a master’s student in the MIT Sloan School of Management. After completing his degree in 2001, he became a consultant to MIT, serving former president Charles Vest and former provost Robert Brown. In 2002, Ruiz formally joined MIT as manager of financial planning and analysis, becoming associate director of the Office of Budget and Financial Planning in 2003. He was named director of finance in 2005, leading an Institute-wide rebalancing that several years later yielded the first balanced general unrestricted budget in 15 years.

Ruiz was named vice president for finance in 2007; in his letter to the community, President Reif noted that MIT has not had an operating loss since then. Ruiz became MIT’s executive vice president and treasurer in 2011.

In addition to strengthening MIT’s financial position, Ruiz has played an instrumental part in many of MIT’s major initiatives of the past decade. For instance, guided by President Reif’s vision of a new model for driving innovation from the lab to commercial reality, Ruiz led the development of The Engine, an accelerator for “tough tech” entrepreneurs from MIT and across the Boston area.

Launched in 2016, The Engine has since provided selected startups with specialized lab infrastructure and substantial capital — as well as an experienced support network. Ruiz brought in significant commitments for The Engine’s first $205 million venture fund and recruited Katie Rae as The Engine’s founding president and CEO. To date, The Engine has supported 20 companies. 

Ruiz’s efforts to foster innovation in Cambridge have also included his extensive work with Provost Martin Schmidt and the MIT Investment Management Company (MITIMCo) to reshape Kendall Square, enhancing the dynamism of what has become home to one of the world’s greatest concentrations of innovative companies. He was central to MIT’s 2017 agreement with the federal government to redevelop the John A. Volpe National Transportation Systems Center, starting a process to turn a 14-acre parcel in Kendall Square into a more vibrant mixed-use site that will benefit MIT’s mission and the Cambridge community.

“Israel’s quiet leadership and trusted partnership have produced a dramatic change in and around Kendall Square,” says MITIMCo President Seth Alexander. “His vision has transformed MIT’s neighborhood into one of the most dynamic innovation districts anywhere in the world.”

In 2018, Ruiz played the key role in developing financial models and funding strategies to underpin the MIT Stephen A. Schwarzman College of Computing, MIT’s most significant structural change since the 1950s. He also played a critical role in negotiating the college’s $350 million foundational gift.

“I feel incredibly fortunate to have worked with Israel on several projects including The Engine and the MIT Schwarzman College of Computing,” says Dean of Engineering Anantha Chandrakasan. “He is highly collaborative, has always engaged energetically with our community in launching new initiatives, and has consistently offered me excellent advice. I continue to be grateful for his extraordinary work and leadership in creating a robust strategy during the 2008 financial crisis, and in keeping MIT on sound financial footing ever since.”

Working alongside outgoing Deputy Executive Vice President Tony Sharon, Ruiz fortified MIT’s administrative and operating units, including Finance, MIT Medical, Human Resources, Environmental Health and Safety, Sustainability, Campus Planning, Facilities, MIT Police, Information Systems and Technology, and Audit. Recognizing the need to fund renewal of the Institute’s aging infrastructure, Ruiz charted the MIT2030 campus framework, which balances new construction, renewal of older buildings, and the development of a staff to maintain these buildings for future generations of faculty and students. He then secured MIT Corporation approval of a $5.2 billion capital plan to bring this vision to life.

“What Israel has led and executed has been amazing to watch,” Chancellor Cynthia Barnhart says. “Through his leadership, vision, and partnership, we are seeing a transformation of our student residences, and of our student life and academic facilities. His deep commitment to our students has been reflected in his sustained support for key elements of the MIT student experience and education, including financial aid, student well-being, MindHandHeart, and MITx.”

Ruiz has also focused on improving services for the broader MIT community. Other accomplishments, working with Executive Director Robin Elices, have included the digital platform MIT Atlas and the MIT Atlas Service Center, as well as a new MIT Welcome Center expected to open in Kendall Square next summer. With Ruiz’s guidance, MIT has also become a leader in providing options and incentives to help employees improve their commutes while reducing vehicles on the road. According to a recent Boston Globe series, “No other major Boston-area employers … offer commuter incentives with the scale and sophistication of MIT’s.”

“I have been inspired by MIT for over two decades, since the very first day I arrived on campus as an admitted student and met Professor Paul Samuelson,” Ruiz says. “I am still mesmerized by what MIT and its brilliant people accomplish for the world.”

de MIT News https://ift.tt/372Puz9

New project advances commitment to expanding graduate housing

MIT is embarking on a project to design and construct new graduate housing at the west end of campus on the site of the West Lot parking area and Building W89 (MIT Police). Currently in an early planning stage, the apartment-style residence hall is expected to provide 550 new graduate student housing beds, completing the October 2017 commitment MIT made to add 950 beds to the graduate housing system on campus. 

Chancellor Cynthia Barnhart, Provost Martin Schmidt, and Executive Vice President and Treasurer Israel Ruiz announced the project in a letter to the graduate student community today. They highlighted the benefits of the project, noting that “graduate students and graduate student families will now have access to the convenience and independence of apartment-style living in a residence hall minutes away from the heart of campus. We are pleased to be diversifying our housing portfolio to include an option that promises to be an exciting alternative to living off-campus for many students.”

New housing will be constructed on West Lot parking site

Construction of a pair of buildings located along Vassar Street, on the site of Building W89 and the West Lot parking area adjacent to Simmons Hall, will begin after the completion of planning and design work that will be conducted over the next couple of years. In a study of potential locations by the Office of Campus Planning, this site stood out as an ideal solution for expanding graduate student housing options while also reinforcing the connection to Fort Washington Park and other graduate student communities, as well as further defining the Vassar Street corridor.

“We are very grateful for the long-term opportunities this project presents,” says Suzy Nelson, vice president and dean for student life. “Giving more graduate students the opportunity to fully engage in the Institute’s vibrant campus life is an essential way of enhancing their MIT experience. This exciting new project makes that possible, and it makes living at MIT more attractive to them and to their families.”

Graduate housing is a top priority

MIT’s graduate student population has been growing steadily over the past two decades. In recent years, two different Graduate Student Housing Working Group reports highlighted student housing as one of MIT’s competitive strengths and noted MIT’s attention to expanding on-campus graduate student housing over the years. In 1980 and 1990, MIT provided housing for 27 percent of its graduate students; today, MIT provides housing for 38 percent of the graduate student population, satisfying the preferences of today’s students, with fewer shared bedrooms and more efficiencies.

Both the working group’s Report to the Provost (May 2014) and the working group’s Report to the Chancellor (August 2018) recommended that MIT undertake a further expansion of student housing, and the findings encouraged MIT to consider flexible and new on-campus housing options. Families, in particular, face housing challenges in the Boston-Cambridge, Massachusetts area based on cost and availability. Both working groups noted that an increase in one- and two-bedroom apartment-style units would help address unmet demand and would give more graduate student families the opportunity to benefit from living on campus.

With these recommendations in mind, MIT made a commitment in October 2017 to its graduate students (and the City of Cambridge, as part of the Volpe zoning petition) to add 950 beds to MIT’s graduate student housing stock. 

Accommodations for student families and singles

As planned, the new West Campus graduate residence will fulfill the 2017 commitment and will respond to the working group reports by providing a mix of housing unit types that align with the evolving needs of graduate students and student families. 

On-campus graduate housing at MIT, which currently includes more than 2,500 beds across eight residence halls, will soon be supplemented by the new graduate student residence hall under construction in Kendall Square, providing more than 450 beds total (replacing Eastgate beds and adding 250 net new beds). The West Campus student residence is expected to add 550 new graduate student housing beds in a range of apartment styles, including single units as well as larger units suitable for families. These initiatives, in concert with the addition of 150 new graduate student housing beds resulting from renovations, will fulfill MIT’s commitment to add 950 beds to its graduate student housing system.

Next steps for the housing project

MIT is planning to work with an experienced third-party campus housing developer to design, develop, and operate the student residence hall. The developer will provide the financial capital for construction, giving MIT the financial flexibility and bandwidth to expedite adding beds while reserving MIT resources for making capital renewal improvements and addressing deferred maintenance work throughout the student housing system. In the upcoming months, MIT and the developer will work together to establish an agreement and formulate a program plan informed by the Graduate Student Housing Working Group reports. 

As part of the project, the MIT Police office, currently in Building W89, is expected to be relocated to accommodate the new complex. Planning for this move is underway, and more details will be shared when available. The project team is also working closely with the Parking and Transportation Office regarding the parking spaces that will be displaced. The Parking and Transportation Office will work with individual parkers on assignments, taking into consideration proximity to housing and offices. 

The graduate community will have opportunities to continue to engage in discussion around the residence project, providing the project team with feedback and valuable input. MIT is committed to begin permitting for this project by the end of December 2020, and construction is expected to begin in 2021-22.

de MIT News https://ift.tt/2PVuwuM

Model beats Wall Street analysts in forecasting business financials

Knowing a company’s true sales can help determine its value. Investors, for instance, often employ financial analysts to predict a company’s upcoming earnings using various public data, computational tools, and their own intuition. Now MIT researchers have developed an automated model that significantly outperforms humans in predicting business sales using very limited, “noisy” data.

In finance, there’s growing interest in using imprecise but frequently generated consumer data — called “alternative data” — to help predict a company’s earnings for trading and investment purposes. Alternative data can comprise credit card purchases, location data from smartphones, or even satellite images showing how many cars are parked in a retailer’s lot. Combining alternative data with more traditional but infrequent ground-truth financial data — such as quarterly earnings, press releases, and stock prices — can paint a clearer picture of a company’s financial health on even a daily or weekly basis.

But, so far, it’s been very difficult to get accurate, frequent estimates using alternative data. In a paper published this week in the Proceedings of ACM Sigmetrics Conference, the researchers describe a model for forecasting financials that uses only anonymized weekly credit card transactions and three-month earning reports.

Tasked with predicting quarterly earnings of more than 30 companies, the model outperformed the combined estimates of expert Wall Street analysts on 57 percent of predictions. Notably, the analysts had access to any available private or public data and other machine-learning models, while the researchers’ model used a very small dataset of the two data types.

“Alternative data are these weird, proxy signals to help track the underlying financials of a company,” says first author Michael Fleder, a postdoc in the Laboratory for Information and Decision Systems (LIDS). “We asked, ‘Can you combine these noisy signals with quarterly numbers to estimate the true financials of a company at high frequencies?’ Turns out the answer is yes.”

The model could give an edge to investors, traders, or companies looking to frequently compare their sales with competitors. Beyond finance, the model could help social and political scientists, for example, to study aggregated, anonymous data on public behavior. “It’ll be useful for anyone who wants to figure out what people are doing,” Fleder says.

Joining Fleder on the paper is EECS Professor Devavrat Shah, who is the director of MIT’s Statistics and Data Science Center, a member of the Laboratory for Information and Decision Systems, a principal investigator for the MIT Institute for Foundations of Data Science, and an adjunct professor at the Tata Institute of Fundamental Research.  

Tackling the “small data” problem

For better or worse, a lot of consumer data is up for sale. Retailers, for instance, can buy credit card transactions or location data to see how many people are shopping at a competitor. Advertisers can use the data to see how their advertisements are impacting sales. But getting those answers still primarily relies on humans. No machine-learning model has been able to adequately crunch the numbers.

Counterintuitively, the problem is actually lack of data. Each financial input, such as a quarterly report or weekly credit card total, is only one number. Quarterly reports over two years total only eight data points. Credit card data for, say, every week over the same period is only roughly another 100 “noisy” data points, meaning they contain potentially uninterpretable information.

“We have a ‘small data’ problem,” Fleder says. “You only get a tiny slice of what people are spending and you have to extrapolate and infer what’s really going on from that fraction of data.”

For their work, the researchers obtained consumer credit card transactions — at typically weekly and biweekly intervals — and quarterly reports for 34 retailers from 2015 to 2018 from a hedge fund. Across all companies, they gathered 306 quarters-worth of data in total.

Computing daily sales is fairly simple in concept. The model assumes a company’s daily sales remain similar, only slightly decreasing or increasing from one day to the next. Mathematically, that means sales values for consecutive days are multiplied by some constant value plus some statistical noise value — which captures some of the inherent randomness in a company’s sales. Tomorrow’s sales, for instance, equal today’s sales multiplied by, say, 0.998 or 1.01, plus the estimated number for noise.

If given accurate model parameters for the daily constant and noise level, a standard inference algorithm can calculate that equation to output an accurate forecast of daily sales. But the trick is calculating those parameters.

Untangling the numbers

That’s where quarterly reports and probability techniques come in handy. In a simple world, a quarterly report could be divided by, say, 90 days to calculate the daily sales (implying sales are roughly constant day-to-day). In reality, sales vary from day to day. Also, including alternative data to help understand how sales vary over a quarter complicates matters: Apart from being noisy, purchased credit card data always consist of some indeterminate fraction of the total sales. All that makes it very difficult to know how exactly the credit card totals factor into the overall sales estimate.

“That requires a bit of untangling the numbers,” Fleder says. “If we observe 1 percent of a company’s weekly sales through credit card transactions, how do we know it’s 1 percent? And, if the credit card data is noisy, how do you know how noisy it is? We don’t have access to the ground truth for daily or weekly sales totals. But the quarterly aggregates help us reason about those totals.”

To do so, the researchers use a variation of the standard inference algorithm, called Kalman filtering or Belief Propagation, which has been used in various technologies from space shuttles to smartphone GPS. Kalman filtering uses data measurements observed over time, containing noise inaccuracies, to generate a probability distribution for unknown variables over a designated timeframe. In the researchers’ work, that means estimating the possible sales of a single day.

To train the model, the technique first breaks down quarterly sales into a set number of measured days, say 90 — allowing sales to vary day-to-day. Then, it matches the observed, noisy credit card data to unknown daily sales. Using the quarterly numbers and some extrapolation, it estimates the fraction of total sales the credit card data likely represents. Then, it calculates each day’s fraction of observed sales, noise level, and an error estimate for how well it made its predictions.

The inference algorithm plugs all those values into the formula to predict daily sales totals. Then, it can sum those totals to get weekly, monthly, or quarterly numbers. Across all 34 companies, the model beat a consensus benchmark — which combines estimates of Wall Street analysts — on 57.2 percent of 306 quarterly predictions.

Next, the researchers are designing the model to analyze a combination of credit card transactions and other alternative data, such as location information. “This isn’t all we can do. This is just a natural starting point,” Fleder says.

de MIT News https://ift.tt/2Z2khcs

A new way to remove contaminants from nuclear wastewater

Nuclear power continues to expand globally, propelled, in part, by the fact that it produces few greenhouse gas emissions while providing steady power output. But along with that expansion comes an increased need for dealing with the large volumes of water used for cooling these plants, which becomes contaminated with radioactive isotopes that require special long-term disposal.

Now, a method developed at MIT provides a way of substantially reducing the volume of contaminated water that needs to be disposed of, instead concentrating the contaminants and allowing the rest of the water to be recycled through the plant’s cooling system. The proposed system is described in the journal Environmental Science and Technology, in a paper by graduate student Mohammad Alkhadra, professor of chemical engineering Martin Bazant, and three others.

The method makes use of a process called shock electrodialysis, which uses an electric field to generate a deionization shockwave in the water. The shockwave pushes the electrically charged particles, or ions, to one side of a tube filled with charged porous material, so that concentrated stream of contaminants can be separated out from the rest of the water. The group discovered that two radionuclide contaminants — isotopes of cobalt and cesium — can be selectively removed from water that also contains boric acid and lithium. After the water stream is cleansed of its cobalt and cesium contaminants, it can be reused in the reactor.

The shock electrodialysis process was initially developed by Bazant and his co-workers as a general method of removing salt from water, as demonstrated in their first scalable prototype four years ago. Now, the team has focused on this more specific application, which could help improve the economics and environmental impact of working nuclear power plants. In ongoing research, they are also continuing to develop a system for removing other contaminants, including lead, from drinking water.

Not only is the new system inexpensive and scalable to large sizes, but in principle it also can deal with a wide range of contaminants, Bazant says. “It’s a single device that can perform a whole range of separations for any specific application,” he says.

In their earlier desalination work, the researchers used measurements of the water’s electrical conductivity to determine how much salt was removed. In the years since then, the team has developed other methods for detecting and quantifying the details of what’s in the concentrated radioactive waste and the cleaned water.

“We carefully measure the composition of all the stuff going in and out,” says Bazant, who is the E.G. Roos Professor of Chemical Engineering as well as a professor of mathematics. “This really opened up a new direction for our research.” They began to focus on separation processes that would be useful for health reasons or that would result in concentrating material that has high value, either for reuse or to offset disposal costs.

The method they developed works for sea water desalination, but it is a relatively energy-intensive process for that application. The energy cost is dramatically lower when the method is used for ion-selective separations from dilute streams such as nuclear plant cooling water. For this application, which also requires expensive disposal, the method makes economic sense, he says. It also hits both of the team’s targets: dealing with high-value materials and helping to safeguard health. The scale of the application is also significant — a single large nuclear plant can circulate about 10 million cubic meters of water per year through its cooling system, Alkhadra says.

For their tests of the system, the researchers used simulated nuclear wastewater based on a recipe provided by Mitsubishi Heavy Industries, which sponsored the research and is a major builder of nuclear plants. In the team’s tests, after a three-stage separation process, they were able to remove 99.5 percent of the cobalt radionuclides in the water while retaining about 43 percent of the water in cleaned-up form so that it could be reused. As much as two-thirds of the water can be reused if the cleanup level is cut back to 98.3 percent of the contaminants removed, the team found.

While the overall method has many potential applications, the nuclear wastewater separation, is “one of the first problems we think we can solve [with this method] that no other solution exists for,” Bazant says. No other practical, continuous, economic method has been found for separating out the radioactive isotopes of cobalt and cesium, the two major contaminants of nuclear wastewater, he adds.

While the method could be used for routine cleanup, it could also make a big difference in dealing with more extreme cases, such as the millions of gallons of contaminated water at the damaged Fukushima Daichi power plant in Japan, where the accumulation of that contaminated water has threatened to overpower the containment systems designed to prevent it from leaking out into the adjacent Pacific. While the new system has so far only been tested at much smaller scales, Bazant says that such large-scale decontamination systems based on this method might be possible “within a few years.”

The research team also included MIT postdocs Kameron Conforti and Tao Gao and graduate student Huanhuan Tian.

de MIT News https://ift.tt/2EvIpuC