More stories

  • in

    How an archeological approach can help leverage biased data in AI to improve medicine

    The classic computer science adage “garbage in, garbage out” lacks nuance when it comes to understanding biased medical data, argue computer science and bioethics professors from MIT, Johns Hopkins University, and the Alan Turing Institute in a new opinion piece published in a recent edition of the New England Journal of Medicine (NEJM). The rising popularity of artificial intelligence has brought increased scrutiny to the matter of biased AI models resulting in algorithmic discrimination, which the White House Office of Science and Technology identified as a key issue in their recent Blueprint for an AI Bill of Rights. 

    When encountering biased data, particularly for AI models used in medical settings, the typical response is to either collect more data from underrepresented groups or generate synthetic data making up for missing parts to ensure that the model performs equally well across an array of patient populations. But the authors argue that this technical approach should be augmented with a sociotechnical perspective that takes both historical and current social factors into account. By doing so, researchers can be more effective in addressing bias in public health. 

    “The three of us had been discussing the ways in which we often treat issues with data from a machine learning perspective as irritations that need to be managed with a technical solution,” recalls co-author Marzyeh Ghassemi, an assistant professor in electrical engineering and computer science and an affiliate of the Abdul Latif Jameel Clinic for Machine Learning in Health (Jameel Clinic), the Computer Science and Artificial Intelligence Laboratory (CSAIL), and Institute of Medical Engineering and Science (IMES). “We had used analogies of data as an artifact that gives a partial view of past practices, or a cracked mirror holding up a reflection. In both cases the information is perhaps not entirely accurate or favorable: Maybe we think that we behave in certain ways as a society — but when you actually look at the data, it tells a different story. We might not like what that story is, but once you unearth an understanding of the past you can move forward and take steps to address poor practices.” 

    Data as artifact 

    In the paper, titled “Considering Biased Data as Informative Artifacts in AI-Assisted Health Care,” Ghassemi, Kadija Ferryman, and Maxine Mackintosh make the case for viewing biased clinical data as “artifacts” in the same way anthropologists or archeologists would view physical objects: pieces of civilization-revealing practices, belief systems, and cultural values — in the case of the paper, specifically those that have led to existing inequities in the health care system. 

    For example, a 2019 study showed that an algorithm widely considered to be an industry standard used health-care expenditures as an indicator of need, leading to the erroneous conclusion that sicker Black patients require the same level of care as healthier white patients. What researchers found was algorithmic discrimination failing to account for unequal access to care.  

    In this instance, rather than viewing biased datasets or lack of data as problems that only require disposal or fixing, Ghassemi and her colleagues recommend the “artifacts” approach as a way to raise awareness around social and historical elements influencing how data are collected and alternative approaches to clinical AI development. 

    “If the goal of your model is deployment in a clinical setting, you should engage a bioethicist or a clinician with appropriate training reasonably early on in problem formulation,” says Ghassemi. “As computer scientists, we often don’t have a complete picture of the different social and historical factors that have gone into creating data that we’ll be using. We need expertise in discerning when models generalized from existing data may not work well for specific subgroups.” 

    When more data can actually harm performance 

    The authors acknowledge that one of the more challenging aspects of implementing an artifact-based approach is being able to assess whether data have been racially corrected: i.e., using white, male bodies as the conventional standard that other bodies are measured against. The opinion piece cites an example from the Chronic Kidney Disease Collaboration in 2021, which developed a new equation to measure kidney function because the old equation had previously been “corrected” under the blanket assumption that Black people have higher muscle mass. Ghassemi says that researchers should be prepared to investigate race-based correction as part of the research process. 

    In another recent paper accepted to this year’s International Conference on Machine Learning co-authored by Ghassemi’s PhD student Vinith Suriyakumar and University of California at San Diego Assistant Professor Berk Ustun, the researchers found that assuming the inclusion of personalized attributes like self-reported race improve the performance of ML models can actually lead to worse risk scores, models, and metrics for minority and minoritized populations.  

    “There’s no single right solution for whether or not to include self-reported race in a clinical risk score. Self-reported race is a social construct that is both a proxy for other information, and deeply proxied itself in other medical data. The solution needs to fit the evidence,” explains Ghassemi. 

    How to move forward 

    This is not to say that biased datasets should be enshrined, or biased algorithms don’t require fixing — quality training data is still key to developing safe, high-performance clinical AI models, and the NEJM piece highlights the role of the National Institutes of Health (NIH) in driving ethical practices.  

    “Generating high-quality, ethically sourced datasets is crucial for enabling the use of next-generation AI technologies that transform how we do research,” NIH acting director Lawrence Tabak stated in a press release when the NIH announced its $130 million Bridge2AI Program last year. Ghassemi agrees, pointing out that the NIH has “prioritized data collection in ethical ways that cover information we have not previously emphasized the value of in human health — such as environmental factors and social determinants. I’m very excited about their prioritization of, and strong investments towards, achieving meaningful health outcomes.” 

    Elaine Nsoesie, an associate professor at the Boston University of Public Health, believes there are many potential benefits to treating biased datasets as artifacts rather than garbage, starting with the focus on context. “Biases present in a dataset collected for lung cancer patients in a hospital in Uganda might be different from a dataset collected in the U.S. for the same patient population,” she explains. “In considering local context, we can train algorithms to better serve specific populations.” Nsoesie says that understanding the historical and contemporary factors shaping a dataset can make it easier to identify discriminatory practices that might be coded in algorithms or systems in ways that are not immediately obvious. She also notes that an artifact-based approach could lead to the development of new policies and structures ensuring that the root causes of bias in a particular dataset are eliminated. 

    “People often tell me that they are very afraid of AI, especially in health. They’ll say, ‘I’m really scared of an AI misdiagnosing me,’ or ‘I’m concerned it will treat me poorly,’” Ghassemi says. “I tell them, you shouldn’t be scared of some hypothetical AI in health tomorrow, you should be scared of what health is right now. If we take a narrow technical view of the data we extract from systems, we could naively replicate poor practices. That’s not the only option — realizing there is a problem is our first step towards a larger opportunity.”  More

  • in

    Helping computer vision and language models understand what they see

    Powerful machine-learning algorithms known as vision and language models, which learn to match text with images, have shown remarkable results when asked to generate captions or summarize videos.

    While these models excel at identifying objects, they often struggle to understand concepts, like object attributes or the arrangement of items in a scene. For instance, a vision and language model might recognize the cup and table in an image, but fail to grasp that the cup is sitting on the table.

    Researchers from MIT, the MIT-IBM Watson AI Lab, and elsewhere have demonstrated a new technique that utilizes computer-generated data to help vision and language models overcome this shortcoming.

    The researchers created a synthetic dataset of images that depict a wide range of scenarios, object arrangements, and human actions, coupled with detailed text descriptions. They used this annotated dataset to “fix” vision and language models so they can learn concepts more effectively. Their technique ensures these models can still make accurate predictions when they see real images.

    When they tested models on concept understanding, the researchers found that their technique boosted accuracy by up to 10 percent. This could improve systems that automatically caption videos or enhance models that provide natural language answers to questions about images, with applications in fields like e-commerce or health care.

    “With this work, we are going beyond nouns in the sense that we are going beyond just the names of objects to more of the semantic concept of an object and everything around it. Our idea was that, when a machine-learning model sees objects in many different arrangements, it will have a better idea of how arrangement matters in a scene,” says Khaled Shehada, a graduate student in the Department of Electrical Engineering and Computer Science and co-author of a paper on this technique.

    Shehada wrote the paper with lead author Paola Cascante-Bonilla, a computer science graduate student at Rice University; Aude Oliva, director of strategic industry engagement at the MIT Schwarzman College of Computing, MIT director of the MIT-IBM Watson AI Lab, and a senior research scientist in the Computer Science and Artificial Intelligence Laboratory (CSAIL); senior author Leonid Karlinsky, a research staff member in the MIT-IBM Watson AI Lab; and others at MIT, the MIT-IBM Watson AI Lab, Georgia Tech, Rice University, École des Ponts, Weizmann Institute of Science, and IBM Research. The paper will be presented at the International Conference on Computer Vision.

    Focusing on objects

    Vision and language models typically learn to identify objects in a scene, and can end up ignoring object attributes, such as color and size, or positional relationships, such as which object is on top of another object.

    This is due to the method with which these models are often trained, known as contrastive learning. This training method involves forcing a model to predict the correspondence between images and text. When comparing natural images, the objects in each scene tend to cause the most striking differences. (Perhaps one image shows a horse in a field while the second shows a sailboat on the water.)

    “Every image could be uniquely defined by the objects in the image. So, when you do contrastive learning, just focusing on the nouns and objects would solve the problem. Why would the model do anything differently?” says Karlinsky.

    The researchers sought to mitigate this problem by using synthetic data to fine-tune a vision and language model. The fine-tuning process involves tweaking a model that has already been trained to improve its performance on a specific task.

    They used a computer to automatically create synthetic videos with diverse 3D environments and objects, such as furniture and luggage, and added human avatars that interacted with the objects.

    Using individual frames of these videos, they generated nearly 800,000 photorealistic images, and then paired each with a detailed caption. The researchers developed a methodology for annotating every aspect of the image to capture object attributes, positional relationships, and human-object interactions clearly and consistently in dense captions.

    Because the researchers created the images, they could control the appearance and position of objects, as well as the gender, clothing, poses, and actions of the human avatars.

    “Synthetic data allows a lot of diversity. With real images, you might not have a lot of elephants in a room, but with synthetic data, you could actually have a pink elephant in a room with a human, if you want,” Cascante-Bonilla says.

    Synthetic data have other advantages, too. They are cheaper to generate than real data, yet the images are highly photorealistic. They also preserve privacy because no real humans are shown in the images. And, because data are produced automatically by a computer, they can be generated quickly in massive quantities.

    By using different camera viewpoints, or slightly changing the positions or attributes of objects, the researchers created a dataset with a far wider variety of scenarios than one would find in a natural dataset.

    Fine-tune, but don’t forget

    However, when one fine-tunes a model with synthetic data, there is a risk that model might “forget” what it learned when it was originally trained with real data.

    The researchers employed a few techniques to prevent this problem, such as adjusting the synthetic data so colors, lighting, and shadows more closely match those found in natural images. They also made adjustments to the model’s inner-workings after fine-tuning to further reduce any forgetfulness.

    Their synthetic dataset and fine-tuning strategy improved the ability of popular vision and language models to accurately recognize concepts by up to 10 percent. At the same time, the models did not forget what they had already learned.

    Now that they have shown how synthetic data can be used to solve this problem, the researchers want to identify ways to improve the visual quality and diversity of these data, as well as the underlying physics that makes synthetic scenes look realistic. In addition, they plan to test the limits of scalability, and investigate whether model improvement starts to plateau with larger and more diverse synthetic datasets.

    This research is funded, in part, by the U.S. Defense Advanced Research Projects Agency, the National Science Foundation, and the MIT-IBM Watson AI Lab. More

  • in

    M’Care and MIT students join forces to improve child health in Nigeria

    Through a collaboration between M’Care, a 2021 Health Security and Pandemics Solver team, and students from MIT, the landscape of child health care in Nigeria could undergo a transformative change, wherein the power of data is harnessed to improve child health outcomes in economically disadvantaged communities. 

    M’Care is a mobile application of Promane and Promade Limited, developed by Opeoluwa Ashimi, which gives community health workers in Nigeria real-time diagnostic and treatment support. The application also creates a dashboard that is available to government health officials to help identify disease trends and deploy timely interventions. As part of its work, M’Care is working to mitigate malnutrition by providing micronutrient powder, vitamin A, and zinc to children below the age of 5. To help deepen its impact, Ashimi decided to work with students in the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) course 6.S897 (Machine Learning for Healthcare) — instructed by professors Peter Szolovits and Manolis Kellis — to leverage data in order to improve nutrient delivery to children across Nigeria. The collaboration also enabled students to see real-world applications for data analysis in the health care space.

    A meeting of minds: M’Care, MIT, and national health authorities

    “Our primary goal for collaborating with the ML for Health team was to spot the missing link in the continuum of care. With over 1 million cumulative consultations that qualify for a continuum of care evaluation, it was important to spot why patients could be lost to followup, prevent this, and ensure completion of care to successfully address the health needs of our patients,” says Ashimi, founder and CEO of M’Care.

    In May 2023, Ashimi attended a meeting that brought together key national stakeholders, including the representatives of the National Ministry of Health in Nigeria. This gathering served as a platform to discuss the profound impact of M’Care’s and ML for Health team’s collaboration — bolstered by data analysis provided on dosage regimens and a child’s age to enhance continuum of care with its attendant impact on children’s health, particularly in relation to brain development with regards to the use of essential micronutrients. The data analyzed by the students using ML methods that were shared during the meeting provided strong supporting evidence to individualize dosage regimens for children based on their age in months for the ANRIN project — a national nutrition project supported by the World Bank — as well as policy decisions to extend months of coverage for children, redefining health care practices in Nigeria.

    MIT students drive change by harnessing the power of data

    At the heart of this collaboration lies the contribution of MIT students. Armed with their dedication and skill in data analysis and machine learning, they played a pivotal role in helping M’Care analyze their data and prepare for their meeting with the Ministry of Health. Their most significant findings included ways to identify patients at risk of not completing their full course of micronutrient powder and/or vitamin A, and identifying gaps in M’Care’s data, such as postdated delivery dates and community demographics. These findings are already helping M’Care better plan its resources and adjust the scope of its program to ensure more children complete the intervention.

    Darcy Kim, an undergraduate at Wellesley College studying math and computer science, who is cross-registered for the MIT machine learning course, expresses enthusiasm about the practical applications found within the project: “To me, data and math is storytelling, and the story is why I love studying it. … I learned that data exploration involves asking questions about how the data is collected, and that surprising patterns that arise often have a qualitative explanation. Impactful research requires radical collaboration with the people the research intends to help. Otherwise, these qualitative explanations get lost in the numbers.”

    Joyce Luo, a first-year operations research PhD student at the Operations Research Center at MIT, shares similar thoughts about the project: “I learned the importance of understanding the context behind data to figure out what kind of analysis might be most impactful. This involves being in frequent contact with the company or organization who provides the data to learn as much as you can about how the data was collected and the people the analysis could help. Stepping back and looking at the bigger picture, rather than just focusing on accuracy or metrics, is extremely important.”

    Insights to implementation: A new era for micronutrient dosing

    As a direct result of M’Care’s collaboration with MIT, policymakers revamped the dosing scheme for essential micronutrient administration for children in Nigeria to prevent malnutrition. M’Care and MIT’s data analysis unearthed critical insights into the limited frequency of medical visits caused by late-age enrollment. 

    “One big takeaway for me was that the data analysis portion of the project — doing a deep dive into the data; understanding, analyzing, visualizing, and summarizing the data — can be just as important as building the machine learning models. M’Care shared our data analysis with the National Ministry of Health, and the insights from it drove them to change their dosing scheme and schedule for delivering micronutrient powder to young children. This really showed us the value of understanding and knowing your data before modeling,” shares Angela Lin, a second-year PhD student at the Operations Research Center.

    Armed with this knowledge, policymakers are eager to develop an optimized dosing scheme that caters to the unique needs of children in disadvantaged communities, ensuring maximum impact on their brain development and overall well-being.

    Siddharth Srivastava, M’Care’s corporate technology liaison, shares his gratitude for the MIT student’s input. “Collaborating with enthusiastic and driven students was both empowering and inspiring. Each of them brought unique perspectives and technical skills to the table. Their passion for applying machine learning to health care was evident in their unwavering dedication and proactive approach to problem-solving.”

    Forging a path to impact

    The collaboration between M’Care and MIT exemplifies the remarkable achievements that arise when academia, innovative problem-solvers, and policy authorities unite. By merging academic rigor with real-world expertise, this partnership has the potential to revolutionize child health care not only in Nigeria but also in similar contexts worldwide.

    “I believe applying innovative methods of machine learning, data gathering, instrumentation, and planning to real problems in the developing world can be highly effective for those countries and highly motivating for our students. I was happy to have such a project in our class portfolio this year and look forward to future opportunities,” says Peter Szolovits, professor of computer science and engineering at MIT.

    By harnessing the power of data, innovation, and collective expertise, this collaboration between M’Care and MIT has the potential to improve equitable child health care in Nigeria. “It has been so fulfilling to see how our team’s work has been able to create even the smallest positive impact in such a short period of time, and it has been amazing to work with a company like Promane and Promade Limited that is so knowledgeable and caring for the communities that they serve,” shares Elizabeth Whittier, a second-year PhD electrical engineering student at MIT. More

  • in

    Artificial intelligence for augmentation and productivity

    The MIT Stephen A. Schwarzman College of Computing has awarded seed grants to seven projects that are exploring how artificial intelligence and human-computer interaction can be leveraged to enhance modern work spaces to achieve better management and higher productivity.

    Funded by Andrew W. Houston ’05 and Dropbox Inc., the projects are intended to be interdisciplinary and bring together researchers from computing, social sciences, and management.

    The seed grants can enable the project teams to conduct research that leads to bigger endeavors in this rapidly evolving area, as well as build community around questions related to AI-augmented management.

    The seven selected projects and research leads include:

    “LLMex: Implementing Vannevar Bush’s Vision of the Memex Using Large Language Models,” led by Patti Maes of the Media Lab and David Karger of the Department of Electrical Engineering and Computer Science (EECS) and the Computer Science and Artificial Intelligence Laboratory (CSAIL). Inspired by Vannevar Bush’s Memex, this project proposes to design, implement, and test the concept of memory prosthetics using large language models (LLMs). The AI-based system will intelligently help an individual keep track of vast amounts of information, accelerate productivity, and reduce errors by automatically recording their work actions and meetings, supporting retrieval based on metadata and vague descriptions, and suggesting relevant, personalized information proactively based on the user’s current focus and context.

    “Using AI Agents to Simulate Social Scenarios,” led by John Horton of the MIT Sloan School of Management and Jacob Andreas of EECS and CSAIL. This project imagines the ability to easily simulate policies, organizational arrangements, and communication tools with AI agents before implementation. Tapping into the capabilities of modern LLMs to serve as a computational model of humans makes this vision of social simulation more realistic, and potentially more predictive.

    “Human Expertise in the Age of AI: Can We Have Our Cake and Eat it Too?” led by Manish Raghavan of MIT Sloan and EECS, and Devavrat Shah of EECS and the Laboratory for Information and Decision Systems. Progress in machine learning, AI, and in algorithmic decision aids has raised the prospect that algorithms may complement human decision-making in a wide variety of settings. Rather than replacing human professionals, this project sees a future where AI and algorithmic decision aids play a role that is complementary to human expertise.

    “Implementing Generative AI in U.S. Hospitals,” led by Julie Shah of the Department of Aeronautics and Astronautics and CSAIL, Retsef Levi of MIT Sloan and the Operations Research Center, Kate Kellog of MIT Sloan, and Ben Armstrong of the Industrial Performance Center. In recent years, studies have linked a rise in burnout from doctors and nurses in the United States with increased administrative burdens associated with electronic health records and other technologies. This project aims to develop a holistic framework to study how generative AI technologies can both increase productivity for organizations and improve job quality for workers in health care settings.

    “Generative AI Augmented Software Tools to Democratize Programming,” led by Harold Abelson of EECS and CSAIL, Cynthia Breazeal of the Media Lab, and Eric Klopfer of the Comparative Media Studies/Writing. Progress in generative AI over the past year is fomenting an upheaval in assumptions about future careers in software and deprecating the role of coding. This project will stimulate a similar transformation in computing education for those who have no prior technical training by creating a software tool that could eliminate much of the need for learners to deal with code when creating applications.

    “Acquiring Expertise and Societal Productivity in a World of Artificial Intelligence,” led by David Atkin and Martin Beraja of the Department of Economics, and Danielle Li of MIT Sloan. Generative AI is thought to augment the capabilities of workers performing cognitive tasks. This project seeks to better understand how the arrival of AI technologies may impact skill acquisition and productivity, and to explore complementary policy interventions that will allow society to maximize the gains from such technologies.

    “AI Augmented Onboarding and Support,” led by Tim Kraska of EECS and CSAIL, and Christoph Paus of the Department of Physics. While LLMs have made enormous leaps forward in recent years and are poised to fundamentally change the way students and professionals learn about new tools and systems, there is often a steep learning curve which people have to climb in order to make full use of the resource. To help mitigate the issue, this project proposes the development of new LLM-powered onboarding and support systems that will positively impact the way support teams operate and improve the user experience. More

  • in

    How machine learning models can amplify inequities in medical diagnosis and treatment

    Prior to receiving a PhD in computer science from MIT in 2017, Marzyeh Ghassemi had already begun to wonder whether the use of AI techniques might enhance the biases that already existed in health care. She was one of the early researchers to take up this issue, and she’s been exploring it ever since. In a new paper, Ghassemi, now an assistant professor in MIT’s Department of Electrical Science and Engineering (EECS), and three collaborators based at the Computer Science and Artificial Intelligence Laboratory, have probed the roots of the disparities that can arise in machine learning, often causing models that perform well overall to falter when it comes to subgroups for which relatively few data have been collected and utilized in the training process. The paper — written by two MIT PhD students, Yuzhe Yang and Haoran Zhang, EECS computer scientist Dina Katabi (the Thuan and Nicole Pham Professor), and Ghassemi — was presented last month at the 40th International Conference on Machine Learning in Honolulu, Hawaii.

    In their analysis, the researchers focused on “subpopulation shifts” — differences in the way machine learning models perform for one subgroup as compared to another. “We want the models to be fair and work equally well for all groups, but instead we consistently observe the presence of shifts among different groups that can lead to inferior medical diagnosis and treatment,” says Yang, who along with Zhang are the two lead authors on the paper. The main point of their inquiry is to determine the kinds of subpopulation shifts that can occur and to uncover the mechanisms behind them so that, ultimately, more equitable models can be developed.

    The new paper “significantly advances our understanding” of the subpopulation shift phenomenon, claims Stanford University computer scientist Sanmi Koyejo. “This research contributes valuable insights for future advancements in machine learning models’ performance on underrepresented subgroups.”

    Camels and cattle

    The MIT group has identified four principal types of shifts — spurious correlations, attribute imbalance, class imbalance, and attribute generalization — which, according to Yang, “have never been put together into a coherent and unified framework. We’ve come up with a single equation that shows you where biases can come from.”

    Biases can, in fact, stem from what the researchers call the class, or from the attribute, or both. To pick a simple example, suppose the task assigned to the machine learning model is to sort images of objects — animals in this case — into two classes: cows and camels. Attributes are descriptors that don’t specifically relate to the class itself. It might turn out, for instance, that all the images used in the analysis show cows standing on grass and camels on sand — grass and sand serving as the attributes here. Given the data available to it, the machine could reach an erroneous conclusion — namely that cows can only be found on grass, not on sand, with the opposite being true for camels. Such a finding would be incorrect, however, giving rise to a spurious correlation, which, Yang explains, is a “special case” among subpopulation shifts — “one in which you have a bias in both the class and the attribute.”

    In a medical setting, one could rely on machine learning models to determine whether a person has pneumonia or not based on an examination of X-ray images. There would be two classes in this situation, one consisting of people who have the lung ailment, another for those who are infection-free. A relatively straightforward case would involve just two attributes: the people getting X-rayed are either female or male. If, in this particular dataset, there were 100 males diagnosed with pneumonia for every one female diagnosed with pneumonia, that could lead to an attribute imbalance, and the model would likely do a better job of correctly detecting pneumonia for a man than for a woman. Similarly, having 1,000 times more healthy (pneumonia-free) subjects than sick ones would lead to a class imbalance, with the model biased toward healthy cases. Attribute generalization is the last shift highlighted in the new study. If your sample contained 100 male patients with pneumonia and zero female subjects with the same illness, you still would like the model to be able to generalize and make predictions about female subjects even though there are no samples in the training data for females with pneumonia.

    The team then took 20 advanced algorithms, designed to carry out classification tasks, and tested them on a dozen datasets to see how they performed across different population groups. They reached some unexpected conclusions: By improving the “classifier,” which is the last layer of the neural network, they were able to reduce the occurrence of spurious correlations and class imbalance, but the other shifts were unaffected. Improvements to the “encoder,” one of the uppermost layers in the neural network, could reduce the problem of attribute imbalance. “However, no matter what we did to the encoder or classifier, we did not see any improvements in terms of attribute generalization,” Yang says, “and we don’t yet know how to address that.”

    Precisely accurate

    There is also the question of assessing how well your model actually works in terms of evenhandedness among different population groups. The metric normally used, called worst-group accuracy or WGA, is based on the assumption that if you can improve the accuracy — of, say, medical diagnosis — for the group that has the worst model performance, you would have improved the model as a whole. “The WGA is considered the gold standard in subpopulation evaluation,” the authors contend, but they made a surprising discovery: boosting worst-group accuracy results in a decrease in what they call “worst-case precision.” In medical decision-making of all sorts, one needs both accuracy — which speaks to the validity of the findings — and precision, which relates to the reliability of the methodology. “Precision and accuracy are both very important metrics in classification tasks, and that is especially true in medical diagnostics,” Yang explains. “You should never trade precision for accuracy. You always need to balance the two.”

    The MIT scientists are putting their theories into practice. In a study they’re conducting with a medical center, they’re looking at public datasets for tens of thousands of patients and hundreds of thousands of chest X-rays, trying to see whether it’s possible for machine learning models to work in an unbiased manner for all populations. That’s still far from the case, even though more awareness has been drawn to this problem, Yang says. “We are finding many disparities across different ages, gender, ethnicity, and intersectional groups.”

    He and his colleagues agree on the eventual goal, which is to achieve fairness in health care among all populations. But before we can reach that point, they maintain, we still need a better understanding of the sources of unfairness and how they permeate our current system. Reforming the system as a whole will not be easy, they acknowledge. In fact, the title of the paper they introduced at the Honolulu conference, “Change is Hard,” gives some indications as to the challenges that they and like-minded researchers face. More

  • in

    A faster way to teach a robot

    Imagine purchasing a robot to perform household tasks. This robot was built and trained in a factory on a certain set of tasks and has never seen the items in your home. When you ask it to pick up a mug from your kitchen table, it might not recognize your mug (perhaps because this mug is painted with an unusual image, say, of MIT’s mascot, Tim the Beaver). So, the robot fails.

    “Right now, the way we train these robots, when they fail, we don’t really know why. So you would just throw up your hands and say, ‘OK, I guess we have to start over.’ A critical component that is missing from this system is enabling the robot to demonstrate why it is failing so the user can give it feedback,” says Andi Peng, an electrical engineering and computer science (EECS) graduate student at MIT.

    Peng and her collaborators at MIT, New York University, and the University of California at Berkeley created a framework that enables humans to quickly teach a robot what they want it to do, with a minimal amount of effort.

    When a robot fails, the system uses an algorithm to generate counterfactual explanations that describe what needed to change for the robot to succeed. For instance, maybe the robot would have been able to pick up the mug if the mug were a certain color. It shows these counterfactuals to the human and asks for feedback on why the robot failed. Then the system utilizes this feedback and the counterfactual explanations to generate new data it uses to fine-tune the robot.

    Fine-tuning involves tweaking a machine-learning model that has already been trained to perform one task, so it can perform a second, similar task.

    The researchers tested this technique in simulations and found that it could teach a robot more efficiently than other methods. The robots trained with this framework performed better, while the training process consumed less of a human’s time.

    This framework could help robots learn faster in new environments without requiring a user to have technical knowledge. In the long run, this could be a step toward enabling general-purpose robots to efficiently perform daily tasks for the elderly or individuals with disabilities in a variety of settings.

    Peng, the lead author, is joined by co-authors Aviv Netanyahu, an EECS graduate student; Mark Ho, an assistant professor at the Stevens Institute of Technology; Tianmin Shu, an MIT postdoc; Andreea Bobu, a graduate student at UC Berkeley; and senior authors Julie Shah, an MIT professor of aeronautics and astronautics and the director of the Interactive Robotics Group in the Computer Science and Artificial Intelligence Laboratory (CSAIL), and Pulkit Agrawal, a professor in CSAIL. The research will be presented at the International Conference on Machine Learning.

    On-the-job training

    Robots often fail due to distribution shift — the robot is presented with objects and spaces it did not see during training, and it doesn’t understand what to do in this new environment.

    One way to retrain a robot for a specific task is imitation learning. The user could demonstrate the correct task to teach the robot what to do. If a user tries to teach a robot to pick up a mug, but demonstrates with a white mug, the robot could learn that all mugs are white. It may then fail to pick up a red, blue, or “Tim-the-Beaver-brown” mug.

    Training a robot to recognize that a mug is a mug, regardless of its color, could take thousands of demonstrations.

    “I don’t want to have to demonstrate with 30,000 mugs. I want to demonstrate with just one mug. But then I need to teach the robot so it recognizes that it can pick up a mug of any color,” Peng says.

    To accomplish this, the researchers’ system determines what specific object the user cares about (a mug) and what elements aren’t important for the task (perhaps the color of the mug doesn’t matter). It uses this information to generate new, synthetic data by changing these “unimportant” visual concepts. This process is known as data augmentation.

    The framework has three steps. First, it shows the task that caused the robot to fail. Then it collects a demonstration from the user of the desired actions and generates counterfactuals by searching over all features in the space that show what needed to change for the robot to succeed.

    The system shows these counterfactuals to the user and asks for feedback to determine which visual concepts do not impact the desired action. Then it uses this human feedback to generate many new augmented demonstrations.

    In this way, the user could demonstrate picking up one mug, but the system would produce demonstrations showing the desired action with thousands of different mugs by altering the color. It uses these data to fine-tune the robot.

    Creating counterfactual explanations and soliciting feedback from the user are critical for the technique to succeed, Peng says.

    From human reasoning to robot reasoning

    Because their work seeks to put the human in the training loop, the researchers tested their technique with human users. They first conducted a study in which they asked people if counterfactual explanations helped them identify elements that could be changed without affecting the task.

    “It was so clear right off the bat. Humans are so good at this type of counterfactual reasoning. And this counterfactual step is what allows human reasoning to be translated into robot reasoning in a way that makes sense,” she says.

    Then they applied their framework to three simulations where robots were tasked with: navigating to a goal object, picking up a key and unlocking a door, and picking up a desired object then placing it on a tabletop. In each instance, their method enabled the robot to learn faster than with other techniques, while requiring fewer demonstrations from users.

    Moving forward, the researchers hope to test this framework on real robots. They also want to focus on reducing the time it takes the system to create new data using generative machine-learning models.

    “We want robots to do what humans do, and we want them to do it in a semantically meaningful way. Humans tend to operate in this abstract space, where they don’t think about every single property in an image. At the end of the day, this is really about enabling a robot to learn a good, human-like representation at an abstract level,” Peng says.

    This research is supported, in part, by a National Science Foundation Graduate Research Fellowship, Open Philanthropy, an Apple AI/ML Fellowship, Hyundai Motor Corporation, the MIT-IBM Watson AI Lab, and the National Science Foundation Institute for Artificial Intelligence and Fundamental Interactions. More

  • in

    A new way to look at data privacy

    Imagine that a team of scientists has developed a machine-learning model that can predict whether a patient has cancer from lung scan images. They want to share this model with hospitals around the world so clinicians can start using it in diagnosis.

    But there’s a problem. To teach their model how to predict cancer, they showed it millions of real lung scan images, a process called training. Those sensitive data, which are now encoded into the inner workings of the model, could potentially be extracted by a malicious agent. The scientists can prevent this by adding noise, or more generic randomness, to the model that makes it harder for an adversary to guess the original data. However, perturbation reduces a model’s accuracy, so the less noise one can add, the better.

    MIT researchers have developed a technique that enables the user to potentially add the smallest amount of noise possible, while still ensuring the sensitive data are protected.

    The researchers created a new privacy metric, which they call Probably Approximately Correct (PAC) Privacy, and built a framework based on this metric that can automatically determine the minimal amount of noise that needs to be added. Moreover, this framework does not need knowledge of the inner workings of a model or its training process, which makes it easier to use for different types of models and applications.

    In several cases, the researchers show that the amount of noise required to protect sensitive data from adversaries is far less with PAC Privacy than with other approaches. This could help engineers create machine-learning models that provably hide training data, while maintaining accuracy in real-world settings.

    “PAC Privacy exploits the uncertainty or entropy of the sensitive data in a meaningful way,  and this allows us to add, in many cases, an order of magnitude less noise. This framework allows us to understand the characteristics of arbitrary data processing and privatize it automatically without artificial modifications. While we are in the early days and we are doing simple examples, we are excited about the promise of this technique,” says Srini Devadas, the Edwin Sibley Webster Professor of Electrical Engineering and co-author of a new paper on PAC Privacy.

    Devadas wrote the paper with lead author Hanshen Xiao, an electrical engineering and computer science graduate student. The research will be presented at the International Cryptography Conference (Crypto 2023).

    Defining privacy

    A fundamental question in data privacy is: How much sensitive data could an adversary recover from a machine-learning model with noise added to it?

    Differential Privacy, one popular privacy definition, says privacy is achieved if an adversary who observes the released model cannot infer whether an arbitrary individual’s data is used for the training processing. But provably preventing an adversary from distinguishing data usage often requires large amounts of noise to obscure it. This noise reduces the model’s accuracy.

    PAC Privacy looks at the problem a bit differently. It characterizes how hard it would be for an adversary to reconstruct any part of randomly sampled or generated sensitive data after noise has been added, rather than only focusing on the distinguishability problem.

    For instance, if the sensitive data are images of human faces, differential privacy would focus on whether the adversary can tell if someone’s face was in the dataset. PAC Privacy, on the other hand, could look at whether an adversary could extract a silhouette — an approximation — that someone could recognize as a particular individual’s face.

    Once they established the definition of PAC Privacy, the researchers created an algorithm that automatically tells the user how much noise to add to a model to prevent an adversary from confidently reconstructing a close approximation of the sensitive data. This algorithm guarantees privacy even if the adversary has infinite computing power, Xiao says.

    To find the optimal amount of noise, the PAC Privacy algorithm relies on the uncertainty, or entropy, in the original data from the viewpoint of the adversary.

    This automatic technique takes samples randomly from a data distribution or a large data pool and runs the user’s machine-learning training algorithm on that subsampled data to produce an output learned model. It does this many times on different subsamplings and compares the variance across all outputs. This variance determines how much noise one must add — a smaller variance means less noise is needed.

    Algorithm advantages

    Different from other privacy approaches, the PAC Privacy algorithm does not need knowledge of the inner workings of a model, or the training process.

    When implementing PAC Privacy, a user can specify their desired level of confidence at the outset. For instance, perhaps the user wants a guarantee that an adversary will not be more than 1 percent confident that they have successfully reconstructed the sensitive data to within 5 percent of its actual value. The PAC Privacy algorithm automatically tells the user the optimal amount of noise that needs to be added to the output model before it is shared publicly, in order to achieve those goals.

    “The noise is optimal, in the sense that if you add less than we tell you, all bets could be off. But the effect of adding noise to neural network parameters is complicated, and we are making no promises on the utility drop the model may experience with the added noise,” Xiao says.

    This points to one limitation of PAC Privacy — the technique does not tell the user how much accuracy the model will lose once the noise is added. PAC Privacy also involves repeatedly training a machine-learning model on many subsamplings of data, so it can be computationally expensive.  

    To improve PAC Privacy, one approach is to modify a user’s machine-learning training process so it is more stable, meaning that the output model it produces does not change very much when the input data is subsampled from a data pool.  This stability would create smaller variances between subsample outputs, so not only would the PAC Privacy algorithm need to be run fewer times to identify the optimal amount of noise, but it would also need to add less noise.

    An added benefit of stabler models is that they often have less generalization error, which means they can make more accurate predictions on previously unseen data, a win-win situation between machine learning and privacy, Devadas adds.

    “In the next few years, we would love to look a little deeper into this relationship between stability and privacy, and the relationship between privacy and generalization error. We are knocking on a door here, but it is not clear yet where the door leads,” he says.

    “Obfuscating the usage of an individual’s data in a model is paramount to protecting their privacy. However, to do so can come at the cost of the datas’ and therefore model’s utility,” says Jeremy Goodsitt, senior machine learning engineer at Capital One, who was not involved with this research. “PAC provides an empirical, black-box solution, which can reduce the added noise compared to current practices while maintaining equivalent privacy guarantees. In addition, its empirical approach broadens its reach to more data consuming applications.”

    This research is funded, in part, by DSTA Singapore, Cisco Systems, Capital One, and a MathWorks Fellowship. More

  • in

    Learning the language of molecules to predict their properties

    Discovering new materials and drugs typically involves a manual, trial-and-error process that can take decades and cost millions of dollars. To streamline this process, scientists often use machine learning to predict molecular properties and narrow down the molecules they need to synthesize and test in the lab.

    Researchers from MIT and the MIT-Watson AI Lab have developed a new, unified framework that can simultaneously predict molecular properties and generate new molecules much more efficiently than these popular deep-learning approaches.

    To teach a machine-learning model to predict a molecule’s biological or mechanical properties, researchers must show it millions of labeled molecular structures — a process known as training. Due to the expense of discovering molecules and the challenges of hand-labeling millions of structures, large training datasets are often hard to come by, which limits the effectiveness of machine-learning approaches.

    By contrast, the system created by the MIT researchers can effectively predict molecular properties using only a small amount of data. Their system has an underlying understanding of the rules that dictate how building blocks combine to produce valid molecules. These rules capture the similarities between molecular structures, which helps the system generate new molecules and predict their properties in a data-efficient manner.

    This method outperformed other machine-learning approaches on both small and large datasets, and was able to accurately predict molecular properties and generate viable molecules when given a dataset with fewer than 100 samples.

    “Our goal with this project is to use some data-driven methods to speed up the discovery of new molecules, so you can train a model to do the prediction without all of these cost-heavy experiments,” says lead author Minghao Guo, a computer science and electrical engineering (EECS) graduate student.

    Guo’s co-authors include MIT-IBM Watson AI Lab research staff members Veronika Thost, Payel Das, and Jie Chen; recent MIT graduates Samuel Song ’23 and Adithya Balachandran ’23; and senior author Wojciech Matusik, a professor of electrical engineering and computer science and a member of the MIT-IBM Watson AI Lab, who leads the Computational Design and Fabrication Group within the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). The research will be presented at the International Conference for Machine Learning.

    Learning the language of molecules

    To achieve the best results with machine-learning models, scientists need training datasets with millions of molecules that have similar properties to those they hope to discover. In reality, these domain-specific datasets are usually very small. So, researchers use models that have been pretrained on large datasets of general molecules, which they apply to a much smaller, targeted dataset. However, because these models haven’t acquired much domain-specific knowledge, they tend to perform poorly.

    The MIT team took a different approach. They created a machine-learning system that automatically learns the “language” of molecules — what is known as a molecular grammar — using only a small, domain-specific dataset. It uses this grammar to construct viable molecules and predict their properties.

    In language theory, one generates words, sentences, or paragraphs based on a set of grammar rules. You can think of a molecular grammar the same way. It is a set of production rules that dictate how to generate molecules or polymers by combining atoms and substructures.

    Just like a language grammar, which can generate a plethora of sentences using the same rules, one molecular grammar can represent a vast number of molecules. Molecules with similar structures use the same grammar production rules, and the system learns to understand these similarities.

    Since structurally similar molecules often have similar properties, the system uses its underlying knowledge of molecular similarity to predict properties of new molecules more efficiently. 

    “Once we have this grammar as a representation for all the different molecules, we can use it to boost the process of property prediction,” Guo says.

    The system learns the production rules for a molecular grammar using reinforcement learning — a trial-and-error process where the model is rewarded for behavior that gets it closer to achieving a goal.

    But because there could be billions of ways to combine atoms and substructures, the process to learn grammar production rules would be too computationally expensive for anything but the tiniest dataset.

    The researchers decoupled the molecular grammar into two parts. The first part, called a metagrammar, is a general, widely applicable grammar they design manually and give the system at the outset. Then it only needs to learn a much smaller, molecule-specific grammar from the domain dataset. This hierarchical approach speeds up the learning process.

    Big results, small datasets

    In experiments, the researchers’ new system simultaneously generated viable molecules and polymers, and predicted their properties more accurately than several popular machine-learning approaches, even when the domain-specific datasets had only a few hundred samples. Some other methods also required a costly pretraining step that the new system avoids.

    The technique was especially effective at predicting physical properties of polymers, such as the glass transition temperature, which is the temperature required for a material to transition from solid to liquid. Obtaining this information manually is often extremely costly because the experiments require extremely high temperatures and pressures.

    To push their approach further, the researchers cut one training set down by more than half — to just 94 samples. Their model still achieved results that were on par with methods trained using the entire dataset.

    “This grammar-based representation is very powerful. And because the grammar itself is a very general representation, it can be deployed to different kinds of graph-form data. We are trying to identify other applications beyond chemistry or material science,” Guo says.

    In the future, they also want to extend their current molecular grammar to include the 3D geometry of molecules and polymers, which is key to understanding the interactions between polymer chains. They are also developing an interface that would show a user the learned grammar production rules and solicit feedback to correct rules that may be wrong, boosting the accuracy of the system.

    This work is funded, in part, by the MIT-IBM Watson AI Lab and its member company, Evonik. More