Technology and society

Subterms

    More stories

    • 88 Shares129 Views

      in Data Management & Statistics

      Nonsense can make sense to machine-learning models

      by

      For all that neural networks can accomplish, we still don’t really understand how they operate. Sure, we can program them to learn, but making sense of a machine’s decision-making process remains much like a fancy puzzle with a dizzying, complex pattern where plenty of integral pieces have yet to be fitted. 

      If a model was trying to classify an image of said puzzle, for example, it could encounter well-known, but annoying adversarial attacks, or even more run-of-the-mill data or processing issues. But a new, more subtle type of failure recently identified by MIT scientists is another cause for concern: “overinterpretation,” where algorithms make confident predictions based on details that don’t make sense to humans, like random patterns or image borders. 

      This could be particularly worrisome for high-stakes environments, like split-second decisions for self-driving cars, and medical diagnostics for diseases that need more immediate attention. Autonomous vehicles in particular rely heavily on systems that can accurately understand surroundings and then make quick, safe decisions. The network used specific backgrounds, edges, or particular patterns of the sky to classify traffic lights and street signs — irrespective of what else was in the image. 

      The team found that neural networks trained on popular datasets like CIFAR-10 and ImageNet suffered from overinterpretation. Models trained on CIFAR-10, for example, made confident predictions even when 95 percent of input images were missing, and the remainder is senseless to humans. 

      “Overinterpretation is a dataset problem that’s caused by these nonsensical signals in datasets. Not only are these high-confidence images unrecognizable, but they contain less than 10 percent of the original image in unimportant areas, such as borders. We found that these images were meaningless to humans, yet models can still classify them with high confidence,” says Brandon Carter, MIT Computer Science and Artificial Intelligence Laboratory PhD student and lead author on a paper about the research. 

      Deep-image classifiers are widely used. In addition to medical diagnosis and boosting autonomous vehicle technology, there are use cases in security, gaming, and even an app that tells you if something is or isn’t a hot dog, because sometimes we need reassurance. The tech in discussion works by processing individual pixels from tons of pre-labeled images for the network to “learn.” 

      Image classification is hard, because machine-learning models have the ability to latch onto these nonsensical subtle signals. Then, when image classifiers are trained on datasets such as ImageNet, they can make seemingly reliable predictions based on those signals. 

      Although these nonsensical signals can lead to model fragility in the real world, the signals are actually valid in the datasets, meaning overinterpretation can’t be diagnosed using typical evaluation methods based on that accuracy. 

      To find the rationale for the model’s prediction on a particular input, the methods in the present study start with the full image and repeatedly ask, what can I remove from this image? Essentially, it keeps covering up the image, until you’re left with the smallest piece that still makes a confident decision. 

      To that end, it could also be possible to use these methods as a type of validation criteria. For example, if you have an autonomously driving car that uses a trained machine-learning method for recognizing stop signs, you could test that method by identifying the smallest input subset that constitutes a stop sign. If that consists of a tree branch, a particular time of day, or something that’s not a stop sign, you could be concerned that the car might come to a stop at a place it’s not supposed to.

      While it may seem that the model is the likely culprit here, the datasets are more likely to blame. “There’s the question of how we can modify the datasets in a way that would enable models to be trained to more closely mimic how a human would think about classifying images and therefore, hopefully, generalize better in these real-world scenarios, like autonomous driving and medical diagnosis, so that the models don’t have this nonsensical behavior,” says Carter. 

      This may mean creating datasets in more controlled environments. Currently, it’s just pictures that are extracted from public domains that are then classified. But if you want to do object identification, for example, it might be necessary to train models with objects with an uninformative background. 

      This work was supported by Schmidt Futures and the National Institutes of Health. Carter wrote the paper alongside Siddhartha Jain and Jonas Mueller, scientists at Amazon, and MIT Professor David Gifford. They are presenting the work at the 2021 Conference on Neural Information Processing Systems. More

    • 150 Shares129 Views

      in Data Management & Statistics

      Systems scientists find clues to why false news snowballs on social media

      by

      The spread of misinformation on social media is a pressing societal problem that tech companies and policymakers continue to grapple with, yet those who study this issue still don’t have a deep understanding of why and how false news spreads.

      To shed some light on this murky topic, researchers at MIT developed a theoretical model of a Twitter-like social network to study how news is shared and explore situations where a non-credible news item will spread more widely than the truth. Agents in the model are driven by a desire to persuade others to take on their point of view: The key assumption in the model is that people bother to share something with their followers if they think it is persuasive and likely to move others closer to their mindset. Otherwise they won’t share.

      The researchers found that in such a setting, when a network is highly connected or the views of its members are sharply polarized, news that is likely to be false will spread more widely and travel deeper into the network than news with higher credibility.

      This theoretical work could inform empirical studies of the relationship between news credibility and the size of its spread, which might help social media companies adapt networks to limit the spread of false information.

      “We show that, even if people are rational in how they decide to share the news, this could still lead to the amplification of information with low credibility. With this persuasion motive, no matter how extreme my beliefs are — given that the more extreme they are the more I gain by moving others’ opinions — there is always someone who would amplify [the information],” says senior author Ali Jadbabaie, professor and head of the Department of Civil and Environmental Engineering and a core faculty member of the Institute for Data, Systems, and Society (IDSS) and a principal investigator in the Laboratory for Information and Decision Systems (LIDS).

      Joining Jadbabaie on the paper are first author Chin-Chia Hsu, a graduate student in the Social and Engineering Systems program in IDSS, and Amir Ajorlou, a LIDS research scientist. The research will be presented this week at the IEEE Conference on Decision and Control.

      Pondering persuasion

      This research draws on a 2018 study by Sinan Aral, the David Austin Professor of Management at the MIT Sloan School of Management; Deb Roy, an associate professor of media arts and sciences at the Media Lab; and former postdoc Soroush Vosoughi (now an assistant professor of computer science at Dartmouth University). Their empirical study of data from Twitter found that false news spreads wider, faster, and deeper than real news.

      Jadbabaie and his collaborators wanted to drill down on why this occurs.

      They hypothesized that persuasion might be a strong motive for sharing news — perhaps agents in the network want to persuade others to take on their point of view — and decided to build a theoretical model that would let them explore this possibility.

      In their model, agents have some prior belief about a policy, and their goal is to persuade followers to move their beliefs closer to the agent’s side of the spectrum.

      A news item is initially released to a small, random subgroup of agents, which must decide whether to share this news with their followers. An agent weighs the newsworthiness of the item and its credibility, and updates its belief based on how surprising or convincing the news is. 

      “They will make a cost-benefit analysis to see if, on average, this piece of news will move people closer to what they think or move them away. And we include a nominal cost for sharing. For instance, taking some action, if you are scrolling on social media, you have to stop to do that. Think of that as a cost. Or a reputation cost might come if I share something that is embarrassing. Everyone has this cost, so the more extreme and the more interesting the news is, the more you want to share it,” Jadbabaie says.

      If the news affirms the agent’s perspective and has persuasive power that outweighs the nominal cost, the agent will always share the news. But if an agent thinks the news item is something others may have already seen, the agent is disincentivized to share it.

      Since an agent’s willingness to share news is a product of its perspective and how persuasive the news is, the more extreme an agent’s perspective or the more surprising the news, the more likely the agent will share it.

      The researchers used this model to study how information spreads during a news cascade, which is an unbroken sharing chain that rapidly permeates the network.

      Connectivity and polarization

      The team found that when a network has high connectivity and the news is surprising, the credibility threshold for starting a news cascade is lower. High connectivity means that there are multiple connections between many users in the network.

      Likewise, when the network is largely polarized, there are plenty of agents with extreme views who want to share the news item, starting a news cascade. In both these instances, news with low credibility creates the largest cascades.

      “For any piece of news, there is a natural network speed limit, a range of connectivity, that facilitates good transmission of information where the size of the cascade is maximized by true news. But if you exceed that speed limit, you will get into situations where inaccurate news or news with low credibility has a larger cascade size,” Jadbabaie says.

      If the views of users in the network become more diverse, it is less likely that a poorly credible piece of news will spread more widely than the truth.

      Jadbabaie and his colleagues designed the agents in the network to behave rationally, so the model would better capture actions real humans might take if they want to persuade others.

      “Someone might say that is not why people share, and that is valid. Why people do certain things is a subject of intense debate in cognitive science, social psychology, neuroscience, economics, and political science,” he says. “Depending on your assumptions, you end up getting different results. But I feel like this assumption of persuasion being the motive is a natural assumption.”

      Their model also shows how costs can be manipulated to reduce the spread of false information. Agents make a cost-benefit analysis and won’t share news if the cost to do so outweighs the benefit of sharing.

      “We don’t make any policy prescriptions, but one thing this work suggests is that, perhaps, having some cost associated with sharing news is not a bad idea. The reason you get lots of these cascades is because the cost of sharing the news is actually very low,” he says.

      This work was supported by an Army Research Office Multidisciplinary University Research Initiative grant and a Vannevar Bush Fellowship from the Office of the Secretary of Defense. More

    • 100 Shares189 Views

      in Data Management & Statistics

      The reasons behind lithium-ion batteries’ rapid cost decline

      by

      Lithium-ion batteries, those marvels of lightweight power that have made possible today’s age of handheld electronics and electric vehicles, have plunged in cost since their introduction three decades ago at a rate similar to the drop in solar panel prices, as documented by a study published last March. But what brought about such an astonishing cost decline, of about 97 percent?

      Some of the researchers behind that earlier study have now analyzed what accounted for the extraordinary savings. They found that by far the biggest factor was work on research and development, particularly in chemistry and materials science. This outweighed the gains achieved through economies of scale, though that turned out to be the second-largest category of reductions.

      The new findings are being published today in the journal Energy and Environmental Science, in a paper by MIT postdoc Micah Ziegler, recent graduate student Juhyun Song PhD ’19, and Jessika Trancik, a professor in MIT’s Institute for Data, Systems and Society.

      The findings could be useful for policymakers and planners to help guide spending priorities in order to continue the pathway toward ever-lower costs for this and other crucial energy storage technologies, according to Trancik. Their work suggests that there is still considerable room for further improvement in electrochemical battery technologies, she says.

      The analysis required digging through a variety of sources, since much of the relevant information consists of closely held proprietary business data. “The data collection effort was extensive,” Ziegler says. “We looked at academic articles, industry and government reports, press releases, and specification sheets. We even looked at some legal filings that came out. We had to piece together data from many different sources to get a sense of what was happening.” He says they collected “about 15,000 qualitative and quantitative data points, across 1,000 individual records from approximately 280 references.”

      Data from the earliest times are hardest to access and can have the greatest uncertainties, Trancik says, but by comparing different data sources from the same period they have attempted to account for these uncertainties.

      Overall, she says, “we estimate that the majority of the cost decline, more than 50 percent, came from research-and-development-related activities.” That included both private sector and government-funded research and development, and “the vast majority” of that cost decline within that R&D category came from chemistry and materials research.

      That was an interesting finding, she says, because “there were so many variables that people were working on through very different kinds of efforts,” including the design of the battery cells themselves, their manufacturing systems, supply chains, and so on. “The cost improvement emerged from a diverse set of efforts and many people, and not from the work of only a few individuals.”

      The findings about the importance of investment in R&D were especially significant, Ziegler says, because much of this investment happened after lithium-ion battery technology was commercialized, a stage at which some analysts thought the research contribution would become less significant. Over roughly a 20-year period starting five years after the batteries’ introduction in the early 1990s, he says, “most of the cost reduction still came from R&D. The R&D contribution didn’t end when commercialization began. In fact, it was still the biggest contributor to cost reduction.”

      The study took advantage of an analytical approach that Trancik and her team initially developed to analyze the similarly precipitous drop in costs of silicon solar panels over the last few decades. They also applied the approach to understand the rising costs of nuclear energy. “This is really getting at the fundamental mechanisms of technological change,” she says. “And we can also develop these models looking forward in time, which allows us to uncover the levers that people could use to improve the technology in the future.”

      One advantage of the methodology Trancik and her colleagues have developed, she says, is that it helps to sort out the relative importance of different factors when many variables are changing all at once, which typically happens as a technology improves. “It’s not simply adding up the cost effects of these variables,” she says, “because many of these variables affect many different cost components. There’s this kind of intricate web of dependencies.” But the team’s methodology makes it possible to “look at how that overall cost change can be attributed to those variables, by essentially mapping out that network of dependencies,” she says.

      This can help provide guidance on public spending, private investments, and other incentives. “What are all the things that different decision makers could do?” she asks. “What decisions do they have agency over so that they could improve the technology, which is important in the case of low-carbon technologies, where we’re looking for solutions to climate change and we have limited time and limited resources? The new approach allows us to potentially be a bit more intentional about where we make those investments of time and money.”

      “This paper collects data available in a systematic way to determine changes in the cost components of lithium-ion batteries between 1990-1995 and 2010-2015,” says Laura Diaz Anadon, a professor of climate change policy at Cambridge University, who was not connected to this research. “This period was an important one in the history of the technology, and understanding the evolution of cost components lays the groundwork for future work on mechanisms and could help inform research efforts in other types of batteries.”

      The research was supported by the Alfred P. Sloan Foundation, the Environmental Defense Fund, and the MIT Technology and Policy Program. More

    • 113 Shares179 Views

      in Data Management & Statistics

      Making machine learning more useful to high-stakes decision makers

      by

      The U.S. Centers for Disease Control and Prevention estimates that one in seven children in the United States experienced abuse or neglect in the past year. Child protective services agencies around the nation receive a high number of reports each year (about 4.4 million in 2019) of alleged neglect or abuse. With so many cases, some agencies are implementing machine learning models to help child welfare specialists screen cases and determine which to recommend for further investigation.

      But these models don’t do any good if the humans they are intended to help don’t understand or trust their outputs.

      Researchers at MIT and elsewhere launched a research project to identify and tackle machine learning usability challenges in child welfare screening. In collaboration with a child welfare department in Colorado, the researchers studied how call screeners assess cases, with and without the help of machine learning predictions. Based on feedback from the call screeners, they designed a visual analytics tool that uses bar graphs to show how specific factors of a case contribute to the predicted risk that a child will be removed from their home within two years.

      The researchers found that screeners are more interested in seeing how each factor, like the child’s age, influences a prediction, rather than understanding the computational basis of how the model works. Their results also show that even a simple model can cause confusion if its features are not described with straightforward language.

      These findings could be applied to other high-risk fields where humans use machine learning models to help them make decisions, but lack data science experience, says senior author Kalyan Veeramachaneni, principal research scientist in the Laboratory for Information and Decision Systems (LIDS) and senior author of the paper.

      “Researchers who study explainable AI, they often try to dig deeper into the model itself to explain what the model did. But a big takeaway from this project is that these domain experts don’t necessarily want to learn what machine learning actually does. They are more interested in understanding why the model is making a different prediction than what their intuition is saying, or what factors it is using to make this prediction. They want information that helps them reconcile their agreements or disagreements with the model, or confirms their intuition,” he says.

      Co-authors include electrical engineering and computer science PhD student Alexandra Zytek, who is the lead author; postdoc Dongyu Liu; and Rhema Vaithianathan, professor of economics and director of the Center for Social Data Analytics at the Auckland University of Technology and professor of social data analytics at the University of Queensland. The research will be presented later this month at the IEEE Visualization Conference.

      Real-world research

      The researchers began the study more than two years ago by identifying seven factors that make a machine learning model less usable, including lack of trust in where predictions come from and disagreements between user opinions and the model’s output.

      With these factors in mind, Zytek and Liu flew to Colorado in the winter of 2019 to learn firsthand from call screeners in a child welfare department. This department is implementing a machine learning system developed by Vaithianathan that generates a risk score for each report, predicting the likelihood the child will be removed from their home. That risk score is based on more than 100 demographic and historic factors, such as the parents’ ages and past court involvements.

      “As you can imagine, just getting a number between one and 20 and being told to integrate this into your workflow can be a bit challenging,” Zytek says.

      They observed how teams of screeners process cases in about 10 minutes and spend most of that time discussing the risk factors associated with the case. That inspired the researchers to develop a case-specific details interface, which shows how each factor influenced the overall risk score using color-coded, horizontal bar graphs that indicate the magnitude of the contribution in a positive or negative direction.

      Based on observations and detailed interviews, the researchers built four additional interfaces that provide explanations of the model, including one that compares a current case to past cases with similar risk scores. Then they ran a series of user studies.

      The studies revealed that more than 90 percent of the screeners found the case-specific details interface to be useful, and it generally increased their trust in the model’s predictions. On the other hand, the screeners did not like the case comparison interface. While the researchers thought this interface would increase trust in the model, screeners were concerned it could lead to decisions based on past cases rather than the current report.   

      “The most interesting result to me was that, the features we showed them — the information that the model uses — had to be really interpretable to start. The model uses more than 100 different features in order to make its prediction, and a lot of those were a bit confusing,” Zytek says.

      Keeping the screeners in the loop throughout the iterative process helped the researchers make decisions about what elements to include in the machine learning explanation tool, called Sibyl.

      As they refined the Sibyl interfaces, the researchers were careful to consider how providing explanations could contribute to some cognitive biases, and even undermine screeners’ trust in the model.

      For instance, since explanations are based on averages in a database of child abuse and neglect cases, having three past abuse referrals may actually decrease the risk score of a child, since averages in this database may be far higher. A screener may see that explanation and decide not to trust the model, even though it is working correctly, Zytek explains. And because humans tend to put more emphasis on recent information, the order in which the factors are listed could also influence decisions.

      Improving interpretability

      Based on feedback from call screeners, the researchers are working to tweak the explanation model so the features that it uses are easier to explain.

      Moving forward, they plan to enhance the interfaces they’ve created based on additional feedback and then run a quantitative user study to track the effects on decision making with real cases. Once those evaluations are complete, they can prepare to deploy Sibyl, Zytek says.

      “It was especially valuable to be able to work so actively with these screeners. We got to really understand the problems they faced. While we saw some reservations on their part, what we saw more of was excitement about how useful these explanations were in certain cases. That was really rewarding,” she says.

      This work is supported, in part, by the National Science Foundation. More

    • 63 Shares159 Views

      in Data Management & Statistics

      Making data visualizations more accessible

      by

      In the early days of the Covid-19 pandemic, the Centers for Disease Control and Prevention produced a simple chart to illustrate how measures like mask wearing and social distancing could “flatten the curve” and reduce the peak of infections.

      The chart was amplified by news sites and shared on social media platforms, but it often lacked a corresponding text description to make it accessible for blind individuals who use a screen reader to navigate the web, shutting out many of the 253 million people worldwide who have visual disabilities.

      This alternative text is often missing from online charts, and even when it is included, it is frequently uninformative or even incorrect, according to qualitative data gathered by scientists at MIT.

      These researchers conducted a study with blind and sighted readers to determine which text is useful to include in a chart description, which text is not, and why. Ultimately, they found that captions for blind readers should focus on the overall trends and statistics in the chart, not its design elements or higher-level insights.

      They also created a conceptual model that can be used to evaluate a chart description, whether the text was generated automatically by software or manually by a human author. Their work could help journalists, academics, and communicators create descriptions that are more effective for blind individuals and guide researchers as they develop better tools to automatically generate captions.

      “Ninety-nine-point-nine percent of images on Twitter lack any kind of description — and that is not hyperbole, that is the actual statistic,” says Alan Lundgard, a graduate student in the Computer Science and Artificial Intelligence Laboratory (CSAIL) and lead author of the paper. “Having people manually author those descriptions seems to be difficult for a variety of reasons. Perhaps semiautonomous tools could help with that. But it is crucial to do this preliminary participatory design work to figure out what is the target for these tools, so we are not generating content that is either not useful to its intended audience or, in the worst case, erroneous.”

      Lundgard wrote the paper with senior author Arvind Satyanarayan, an assistant professor of computer science who leads the Visualization Group in CSAIL. The research will be presented at the Institute of Electrical and Electronics Engineers Visualization Conference in October.

      Evaluating visualizations

      To develop the conceptual model, the researchers planned to begin by studying graphs featured by popular online publications such as FiveThirtyEight and NYTimes.com, but they ran into a problem — those charts mostly lacked any textual descriptions. So instead, they collected descriptions for these charts from graduate students in an MIT data visualization class and through an online survey, then grouped the captions into four categories.

      Level 1 descriptions focus on the elements of the chart, such as its title, legend, and colors. Level 2 descriptions describe statistical content, like the minimum, maximum, or correlations. Level 3 descriptions cover perceptual interpretations of the data, like complex trends or clusters. Level 4 descriptions include subjective interpretations that go beyond the data and draw on the author’s knowledge.

      In a study with blind and sighted readers, the researchers presented visualizations with descriptions at different levels and asked participants to rate how useful they were. While both groups agreed that level 1 content on its own was not very helpful, sighted readers gave level 4 content the highest marks while blind readers ranked that content among the least useful.

      Survey results revealed that a majority of blind readers were emphatic that descriptions should not contain an author’s editorialization, but rather stick to straight facts about the data. On the other hand, most sighted readers preferred a description that told a story about the data.

      “For me, a surprising finding about the lack of utility for the highest-level content is that it ties very closely to feelings about agency and control as a disabled person. In our research, blind readers specifically didn’t want the descriptions to tell them what to think about the data. They want the data to be accessible in a way that allows them to interpret it for themselves, and they want to have the agency to do that interpretation,” Lundgard says.

      A more inclusive future

      This work could have implications as data scientists continue to develop and refine machine learning methods for autogenerating captions and alternative text.

      “We are not able to do it yet, but it is not inconceivable to imagine that in the future we would be able to automate the creation of some of this higher-level content and build models that target level 2 or level 3 in our framework. And now we know what the research questions are. If we want to produce these automated captions, what should those captions say? We are able to be a bit more directed in our future research because we have these four levels,” Satyanarayan says.

      In the future, the four-level framework could also help researchers develop machine learning models that can automatically suggest effective visualizations as part of the data analysis process, or models that can extract the most useful information from a chart.

      This research could also inform future work in Satyanarayan’s group that seeks to make interactive visualizations more accessible for blind readers who use a screen reader to access and interpret the information. 

      “The question of how to ensure that charts and graphs are accessible to screen reader users is both a socially important equity issue and a challenge that can advance the state-of-the-art in AI,” says Meredith Ringel Morris, director and principal scientist of the People + AI Research team at Google Research, who was not involved with this study. “By introducing a framework for conceptualizing natural language descriptions of information graphics that is grounded in end-user needs, this work helps ensure that future AI researchers will focus their efforts on problems aligned with end-users’ values.”

      Morris adds: “Rich natural-language descriptions of data graphics will not only expand access to critical information for people who are blind, but will also benefit a much wider audience as eyes-free interactions via smart speakers, chatbots, and other AI-powered agents become increasingly commonplace.”

      This research was supported by the National Science Foundation. More

    • 100 Shares149 Views

      in Data Management & Statistics

      Data flow’s decisive role on the global stage

      by

      In 2016, Meicen Sun came to a profound realization: “The control of digital information will lie at the heart of all the big questions and big contentions in politics.” A graduate student in her final year of study who is specializing in international security and the political economy of technology, Sun vividly recalls the emergence of the internet “as a democratizing force, an opener, an equalizer,” helping give rise to the Arab Spring. But she was also profoundly struck when nations in the Middle East and elsewhere curbed internet access to throttle citizens’ efforts to speak and mobilize freely.

      During her undergraduate and graduate studies, which came to focus on China and its expanding global role, Sun became convinced that digital constraints initially intended to prevent the free flow of ideas were also having enormous and growing economic impacts.

      “With an exceptionally high mobile internet adoption rate and the explosion of indigenous digital apps, China’s digital economy was surging, helping to drive the nation’s broader economic growth and international competitiveness,” Sun says. “Yet at the same time, the country maintained the most tightly controlled internet ecosystem in the world.”

      Sun set out to explore this apparent paradox in her dissertation. Her research to date has yielded both novel findings and troubling questions.  

      “Through its control of the internet, China has in effect provided protectionist benefits to its own data-intensive domestic sectors,” she says. “If there is a benefit to imposing internet control, given the absence of effective international regulations, does this give authoritarian states an advantage in trade and national competitiveness?” Following this thread, Sun asks, “What might this mean for the future of democracy as the world grows increasingly dependent on digital technology?”

      Protect or innovate

      Early in her graduate program, classes in capitalism and technology and public policy, says Sun, “cemented for me the idea of data as a factor of production, and the importance of cross-border information flow in making a country innovative.” This central premise serves as a springboard for Sun’s doctoral studies.

      In a series of interconnected research papers using China as her primary case, she is examining the double-edged nature of internet limits. “They accord protectionist benefits to domestic data-internet-intensive sectors, on the one hand, but on the other, act as a potential longer-term deterrent to the country’s capacity to innovate.”

      To pursue her doctoral project, advised by professor of political science Kenneth Oye, Sun is extracting data from a multitude of sources, including a website that has been routinely testing web domain accessibility from within China since 2011. This allows her to pin down when and to what degree internet control occurs. She can then compare this information to publicly available records on the expansion or contraction of data-intensive industrial sectors, enabling her to correlate internet control to a sector’s performance.

      Sun has also compiled datasets for firm-level revenue, scientific citations, and patents that permit her to measure aspects of China’s innovation culture. In analyzing her data she leverages both quantitative and qualitative methods, including one co-developed by her dissertation co-advisor, associate professor of political science In Song Kim. Her initial analysis suggests internet control prevents scholars from accessing knowledge available on foreign websites, and that if sustained, such control could take a toll on the Chinese economy over time.

      Of particular concern is the possibility that the economic success that flows from strict internet controls, as exemplified by the Chinese model, may encourage the rise of similar practices among emerging states or those in political flux.

      “The grim implication of my research is that without international regulation on information flow restrictions, democracies will be at a disadvantage against autocracies,” she says. “No matter how short-term or narrow these curbs are, they confer concrete benefits on certain economic sectors.”

      Data, politics, and economy

      Sun got a quick start as a student of China and its role in the world. She was born in Xiamen, a coastal Chinese city across from Taiwan, to academic parents who cultivated her interest in international politics. “My dad would constantly talk to me about global affairs, and he was passionate about foreign policy,” says Sun.

      Eager for education and a broader view of the world, Sun took a scholarship at 15 to attend school in Singapore. “While this experience exposed me to a variety of new ideas and social customs, I felt the itch to travel even farther away, and to meet people with different backgrounds and viewpoints from mine,” than she says.

      Sun attended Princeton University where, after two years sticking to her “comfort zone” — writing and directing plays and composing music for them — she underwent a process of intellectual transition. Political science classes opened a window onto a larger landscape to which she had long been connected: China’s behavior as a rising power and the shifting global landscape.

      She completed her undergraduate degree in politics, and followed up with a master’s degree in international relations at the University of Pennsylvania, where she focused on China-U.S. relations and China’s participation in international institutions. She was on the path to completing a PhD at Penn when, Sun says, “I became confident in my perception that digital technology, and especially information sharing, were becoming critically important factors in international politics, and I felt a strong desire to devote my graduate studies, and even my career, to studying these topics,”

      Certain that the questions she hoped to pursue could best be addressed through an interdisciplinary approach with those working on similar issues, Sun began her doctoral program anew at MIT.

      “Doer mindset”

      Sun is hopeful that her doctoral research will prove useful to governments, policymakers, and business leaders. “There are a lot of developing states actively shopping between data governance and development models for their own countries,” she says. “My findings around the pros and cons of information flow restrictions should be of interest to leaders in these places, and to trade negotiators and others dealing with the global governance of data and what a fair playing field for digital trade would be.”

      Sun has engaged directly with policy and industry experts through her fellowships with the World Economic Forum and the Pacific Forum. And she has embraced questions that touch on policy outside of her immediate research: Sun is collaborating with her dissertation co-advisor, MIT Sloan Professor Yasheng Huang, on a study of the political economy of artificial intelligence in China for the MIT Task Force on the Work of the Future.

      This year, as she writes her dissertation papers, Sun will be based at Georgetown University, where she has a Mortara Center Global Political Economy Project Predoctoral Fellowship. In Washington, she will continue her journey to becoming a “policy-minded scholar, a thinker with a doer mindset, whose findings have bearing on things that happen in the world.” More

    • 63 Shares149 Views

      in Data Management & Statistics

      How quickly do algorithms improve?

      by

      Algorithms are sort of like a parent to a computer. They tell the computer how to make sense of information so they can, in turn, make something useful out of it.

      The more efficient the algorithm, the less work the computer has to do. For all of the technological progress in computing hardware, and the much debated lifespan of Moore’s Law, computer performance is only one side of the picture.

      Behind the scenes a second trend is happening: Algorithms are being improved, so in turn less computing power is needed. While algorithmic efficiency may have less of a spotlight, you’d definitely notice if your trusty search engine suddenly became one-tenth as fast, or if moving through big datasets felt like wading through sludge.

      This led scientists from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) to ask: How quickly do algorithms improve?  

      Existing data on this question were largely anecdotal, consisting of case studies of particular algorithms that were assumed to be representative of the broader scope. Faced with this dearth of evidence, the team set off to crunch data from 57 textbooks and more than 1,110 research papers, to trace the history of when algorithms got better. Some of the research papers directly reported how good new algorithms were, and others needed to be reconstructed by the authors using “pseudocode,” shorthand versions of the algorithm that describe the basic details.

      In total, the team looked at 113 “algorithm families,” sets of algorithms solving the same problem that had been highlighted as most important by computer science textbooks. For each of the 113, the team reconstructed its history, tracking each time a new algorithm was proposed for the problem and making special note of those that were more efficient. Ranging in performance and separated by decades, starting from the 1940s to now, the team found an average of eight algorithms per family, of which a couple improved its efficiency. To share this assembled database of knowledge, the team also created Algorithm-Wiki.org.

      The scientists charted how quickly these families had improved, focusing on the most-analyzed feature of the algorithms — how fast they could guarantee to solve the problem (in computer speak: “worst-case time complexity”). What emerged was enormous variability, but also important insights on how transformative algorithmic improvement has been for computer science.

      For large computing problems, 43 percent of algorithm families had year-on-year improvements that were equal to or larger than the much-touted gains from Moore’s Law. In 14 percent of problems, the improvement to performance from algorithms vastly outpaced those that have come from improved hardware. The gains from algorithm improvement were particularly large for big-data problems, so the importance of those advancements has grown in recent decades.

      The single biggest change that the authors observed came when an algorithm family transitioned from exponential to polynomial complexity. The amount of effort it takes to solve an exponential problem is like a person trying to guess a combination on a lock. If you only have a single 10-digit dial, the task is easy. With four dials like a bicycle lock, it’s hard enough that no one steals your bike, but still conceivable that you could try every combination. With 50, it’s almost impossible — it would take too many steps. Problems that have exponential complexity are like that for computers: As they get bigger they quickly outpace the ability of the computer to handle them. Finding a polynomial algorithm often solves that, making it possible to tackle problems in a way that no amount of hardware improvement can.

      As rumblings of Moore’s Law coming to an end rapidly permeate global conversations, the researchers say that computing users will increasingly need to turn to areas like algorithms for performance improvements. The team says the findings confirm that historically, the gains from algorithms have been enormous, so the potential is there. But if gains come from algorithms instead of hardware, they’ll look different. Hardware improvement from Moore’s Law happens smoothly over time, and for algorithms the gains come in steps that are usually large but infrequent. 

      “This is the first paper to show how fast algorithms are improving across a broad range of examples,” says Neil Thompson, an MIT research scientist at CSAIL and the Sloan School of Management and senior author on the new paper. “Through our analysis, we were able to say how many more tasks could be done using the same amount of computing power after an algorithm improved. As problems increase to billions or trillions of data points, algorithmic improvement becomes substantially more important than hardware improvement. In an era where the environmental footprint of computing is increasingly worrisome, this is a way to improve businesses and other organizations without the downside.”

      Thompson wrote the paper alongside MIT visiting student Yash Sherry. The paper is published in the Proceedings of the IEEE. The work was funded by the Tides foundation and the MIT Initiative on the Digital Economy. More

    • 138 Shares189 Views

      in Data Management & Statistics

      Exact symbolic artificial intelligence for faster, better assessment of AI fairness

      by

      The justice system, banks, and private companies use algorithms to make decisions that have profound impacts on people’s lives. Unfortunately, those algorithms are sometimes biased — disproportionately impacting people of color as well as individuals in lower income classes when they apply for loans or jobs, or even when courts decide what bail should be set while a person awaits trial.

      MIT researchers have developed a new artificial intelligence programming language that can assess the fairness of algorithms more exactly, and more quickly, than available alternatives.

      Their Sum-Product Probabilistic Language (SPPL) is a probabilistic programming system. Probabilistic programming is an emerging field at the intersection of programming languages and artificial intelligence that aims to make AI systems much easier to develop, with early successes in computer vision, common-sense data cleaning, and automated data modeling. Probabilistic programming languages make it much easier for programmers to define probabilistic models and carry out probabilistic inference — that is, work backward to infer probable explanations for observed data.

      “There are previous systems that can solve various fairness questions. Our system is not the first; but because our system is specialized and optimized for a certain class of models, it can deliver solutions thousands of times faster,” says Feras Saad, a PhD student in electrical engineering and computer science (EECS) and first author on a recent paper describing the work. Saad adds that the speedups are not insignificant: The system can be up to 3,000 times faster than previous approaches.

      SPPL gives fast, exact solutions to probabilistic inference questions such as “How likely is the model to recommend a loan to someone over age 40?” or “Generate 1,000 synthetic loan applicants, all under age 30, whose loans will be approved.” These inference results are based on SPPL programs that encode probabilistic models of what kinds of applicants are likely, a priori, and also how to classify them. Fairness questions that SPPL can answer include “Is there a difference between the probability of recommending a loan to an immigrant and nonimmigrant applicant with the same socioeconomic status?” or “What’s the probability of a hire, given that the candidate is qualified for the job and from an underrepresented group?”

      SPPL is different from most probabilistic programming languages, as SPPL only allows users to write probabilistic programs for which it can automatically deliver exact probabilistic inference results. SPPL also makes it possible for users to check how fast inference will be, and therefore avoid writing slow programs. In contrast, other probabilistic programming languages such as Gen and Pyro allow users to write down probabilistic programs where the only known ways to do inference are approximate — that is, the results include errors whose nature and magnitude can be hard to characterize.

      Error from approximate probabilistic inference is tolerable in many AI applications. But it is undesirable to have inference errors corrupting results in socially impactful applications of AI, such as automated decision-making, and especially in fairness analysis.

      Jean-Baptiste Tristan, associate professor at Boston College and former research scientist at Oracle Labs, who was not involved in the new research, says, “I’ve worked on fairness analysis in academia and in real-world, large-scale industry settings. SPPL offers improved flexibility and trustworthiness over other PPLs on this challenging and important class of problems due to the expressiveness of the language, its precise and simple semantics, and the speed and soundness of the exact symbolic inference engine.”

      SPPL avoids errors by restricting to a carefully designed class of models that still includes a broad class of AI algorithms, including the decision tree classifiers that are widely used for algorithmic decision-making. SPPL works by compiling probabilistic programs into a specialized data structure called a “sum-product expression.” SPPL further builds on the emerging theme of using probabilistic circuits as a representation that enables efficient probabilistic inference. This approach extends prior work on sum-product networks to models and queries expressed via a probabilistic programming language. However, Saad notes that this approach comes with limitations: “SPPL is substantially faster for analyzing the fairness of a decision tree, for example, but it can’t analyze models like neural networks. Other systems can analyze both neural networks and decision trees, but they tend to be slower and give inexact answers.”

      “SPPL shows that exact probabilistic inference is practical, not just theoretically possible, for a broad class of probabilistic programs,” says Vikash Mansinghka, an MIT principal research scientist and senior author on the paper. “In my lab, we’ve seen symbolic inference driving speed and accuracy improvements in other inference tasks that we previously approached via approximate Monte Carlo and deep learning algorithms. We’ve also been applying SPPL to probabilistic programs learned from real-world databases, to quantify the probability of rare events, generate synthetic proxy data given constraints, and automatically screen data for probable anomalies.”

      The new SPPL probabilistic programming language was presented in June at the ACM SIGPLAN International Conference on Programming Language Design and Implementation (PLDI), in a paper that Saad co-authored with MIT EECS Professor Martin Rinard and Mansinghka. SPPL is implemented in Python and is available open source. More