CINXE.COM
Anqi (Angie) Liu
<!DOCTYPE html> <html lang="en-us"> <head> <meta charset="utf-8"> <meta name="viewport" content="width=device-width, initial-scale=1"> <meta http-equiv="X-UA-Compatible" content="IE=edge"> <meta name="generator" content="Source Themes Academic 2.4.0"> <meta name="generator" content="Hugo 0.50" /> <meta name="author" content="Anqi (Angie) Liu"> <meta name="description" content="Assistant Professor"> <link rel="alternate" hreflang="en-us" href="https://anqiliu-ai.github.io/"> <meta name="theme-color" content="#0095eb"> <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/4.1.3/css/bootstrap.min.css" integrity="sha256-eSi1q2PG6J7g7ib17yAaWMcrr5GrtohYChqibrV7PBE=" crossorigin="anonymous"> <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/academicons/1.8.6/css/academicons.min.css" integrity="sha256-uFVgMKfistnJAfoCUQigIl+JfUaP47GrRKjf6CTPVmw=" crossorigin="anonymous"> <link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.3.1/css/all.css" integrity="sha384-mzrmE5qonljUremFsqc01SB46JvROS7bZs3IO2EmfFsd15uHvIt+Y8vEf7N7fWAU" crossorigin="anonymous"> <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/fancybox/3.2.5/jquery.fancybox.min.css" integrity="sha256-ygkqlh3CYSUri3LhQxzdcm0n1EQvH2Y+U5S2idbLtxs=" crossorigin="anonymous"> <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/9.12.0/styles/github.min.css" crossorigin="anonymous"> <link rel="stylesheet" href=//fonts.googleapis.com/css?family=Averia+Serif+Libre|Merriweather|Roboto+Mono> <link rel="stylesheet" href="/styles.css"> <link rel="alternate" href="https://anqiliu-ai.github.io/index.xml" type="application/rss+xml" title="Anqi (Angie) Liu"> <link rel="feed" href="https://anqiliu-ai.github.io/index.xml" type="application/rss+xml" title="Anqi (Angie) Liu"> <link rel="manifest" href="/site.webmanifest"> <link rel="icon" type="image/png" href="/img/icon.png"> <link rel="apple-touch-icon" type="image/png" href="/img/icon-192.png"> <link rel="canonical" href="https://anqiliu-ai.github.io/"> <meta property="twitter:card" content="summary_large_image"> <meta property="og:site_name" content="Anqi (Angie) Liu"> <meta property="og:url" content="https://anqiliu-ai.github.io/"> <meta property="og:title" content="Anqi (Angie) Liu"> <meta property="og:description" content="Assistant Professor"> <meta property="og:image" content="https://anqiliu-ai.github.io/img/icon-192.png"> <meta property="og:locale" content="en-us"> <meta property="og:updated_time" content="2022-07-05T00:00:00+00:00"> <title>Anqi (Angie) Liu</title> </head> <body id="top" data-spy="scroll" data-target="#navbar-main" data-offset="71" > <nav class="navbar navbar-light fixed-top navbar-expand-lg py-0" id="navbar-main"> <div class="container"> <a class="navbar-brand" href="/">Anqi (Angie) Liu</a> <button type="button" class="navbar-toggler" data-toggle="collapse" data-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label=""> <span><i class="fas fa-bars"></i></span> </button> <div class="collapse navbar-collapse" id="navbar"> <ul class="navbar-nav ml-auto"> <li class="nav-item"> <a class="nav-link" href="/#about" data-target="#about"> <span>Home</span> </a> </li> <li class="nav-item"> <a class="nav-link" href="/#projects" data-target="#projects"> <span>Projects</span> </a> </li> <li class="nav-item"> <a class="nav-link" href="/publications/" data-target="publications/"> <span>Publications</span> </a> </li> <li class="nav-item"> <a class="nav-link" href="/students/" data-target="students/"> <span>Students</span> </a> </li> <li class="nav-item"> <a class="nav-link" href="/teaching/" data-target="teaching/"> <span>Teaching</span> </a> </li> <li class="nav-item"> <a class="nav-link" href="/#contact" data-target="#contact"> <span>Contact</span> </a> </li> </ul> </div> </div> </nav> <span id="homepage" style="display: none"></span> <section id="about" class="home-section"> <div class="container"> <div class="row" itemprop="author" itemscope itemtype="http://schema.org/Person" itemref="person-email person-address"> <div class="col-12 col-lg-4"> <div id="profile"> <img class="portrait" src="/img/profile.png" itemprop="image"> <div class="portrait-title"> <h2 itemprop="name">Anqi (Angie) Liu</h2> <h3 itemprop="jobTitle">Assistant Professor</h3> <h3 itemprop="worksFor" itemscope itemtype="http://schema.org/Organization"> <a href="https://www.cs.jhu.edu/" target="_blank" itemprop="url" rel="noopener"> <span itemprop="name">CS Department</span> </a> </h3> <h3 itemprop="worksFor" itemscope itemtype="http://schema.org/Organization"> <a href="https://engineering.jhu.edu/" target="_blank" itemprop="url" rel="noopener"> <span itemprop="name">Whiting School of Engineering</span> </a> </h3> <h3 itemprop="worksFor" itemscope itemtype="http://schema.org/Organization"> <a href="https://www.jhu.edu/" target="_blank" itemprop="url" rel="noopener"> <span itemprop="name">Johns Hopkins University</span> </a> </h3> </div> <link itemprop="url" href="https://anqiliu-ai.github.io/"> <ul class="network-icon" aria-hidden="true"> <li> <a itemprop="sameAs" href="mailto:aliu@cs.jhu.edu" target="_blank" rel="noopener"> <i class="fas fa-envelope big-icon"></i> </a> </li> <li> <a itemprop="sameAs" href="https://twitter.com/anqi_liu33" target="_blank" rel="noopener"> <i class="fab fa-twitter big-icon"></i> </a> </li> <li> <a itemprop="sameAs" href="https://scholar.google.com/citations?user=Q8yp6zQAAAAJ&hl=en" target="_blank" rel="noopener"> <i class="ai ai-google-scholar big-icon"></i> </a> </li> </ul> </div> </div> <div class="col-12 col-lg-8" itemprop="description"> <p>I am an Assistant Professor in the <a href="https://www.cs.jhu.edu/" target="_blank">CS department</a> at the <a href="https://engineering.jhu.edu/" target="_blank">Whiting School of Engineering</a> of the <a href="https://www.jhu.edu/" target="_blank">Johns Hopkins University</a>. I am also affiliated with the <a href="https://www.minds.jhu.edu/" target="_blank">Johns Hopkins Mathematical Institute for Data Science (MINDS)</a> and the <a href="https://iaa.jhu.edu/" target="_blank">Johns Hopkins Institute for Assured Autonomy (IAA)</a>. I collaborate extensitively with the <a href="https://www.clsp.jhu.edu/" target="_blank">Center for Language and Speech Processing (CLSP)</a> and the <a href="https://lcsr.jhu.edu/" target="_blank">Laboratory for Computational Sensing and Robotics (LCSR)</a>.</p> <!-- I am looking for motivated students to join my group. Details [here](https://anqiliu-ai.github.io/students/). --> <p>My research interest lies in machine learning for trustworthy AI. I am broadly interested in developing principled machine learning algorithms for building more reliable, trustworthy, and human-compatible AI systems in the real world. This requires the machine learning algorithms to be robust to the changing data and environments, to provide accurate and honest uncertainty estimates, and to consider human preferences and values in the interaction. I am particularly interested in high-stake applications that concern the safety and societal impact of AI.</p> <p>I develop, analyze, and apply methods in statistical machine learning, deep learning, and sequential decision making. One established line of work is in <a href="https://anqiliu-ai.github.io/project/drl/" target="_blank">distributionally robust learning under covariate shift</a>. My recent projects cover topics in different types of distribution shift, active learning, safe exploration, off-policy learning, fair machine learning.</p> <p>I worked with <a href="http://www.yisongyue.com/index.php" target="_blank">Prof. Yisong Yue</a> and <a href="http://tensorlab.cms.caltech.edu/users/anima/" target="_blank">Prof. Anima Anandkumar</a> as a postdoc in the Department of Computing and Mathematical Sciences (<a href="http://www.cms.caltech.edu/" target="_blank">CMS</a>) of California Institute of Technology (<a href="http://www.caltech.edu/" target="_blank">Caltech</a>). Before that, I received my Ph.D. in <a href="https://www.cs.uic.edu/" target="_blank">Department of Computer Science</a>, University of Illinois at Chicago (<a href="https://uic.edu/" target="_blank">UIC</a>). I was very fortunate to have <a href="https://www.cs.uic.edu/Ziebart" target="_blank">Prof. Brian Ziebart</a> as my advisor.</p> <!-- <span style="font-size:larger;">News:</space> --> <!-- Hao Liu, Anqi Liu, Tongxin Li, Anima Anandkumar "Disentangling Causal Effects from Latent Confounders using Interventions", in NeurIPS Workshop on Causal Machine Learning, 2019. PDF coming soon.Anqi Liu, Maya Srikanth, Nicholas Adams-Cohen, R. Michael Alvarez, and Anima Anandkumar "[Finding Social Media Trolls: Dynamic Keyword Selection Methods for Rapidly-Evolving Online Debates](https://arxiv.org/abs/1911.05332)", in NeurIPS Joint Workshop on AI for Social Good, 2019.Quanying Liu, Haiyan Wu, Anqi Liu "[Modeling and Interpreting Real-world Human Risk Decision Making with Inverse Reinforcement Learning](https://arxiv.org/abs/1906.05803)", in Real-world Sequential Decision Making Workshop at ICML 2019. --> <!-- <span style="font-size:larger;">Project Highlights:</space> Distributionally Robust Learning under Covariate Shift (Fundamentals Explained, from IID to Shift) Safe Exploration in Control and Robotics Label-Efficient Learning under Distribution Shift AI for Social Science (Computational Social Science) --> <p><br><br> <span style="font-size:larger;">Selected Recent News:</space></p> <p>Received the <a href="https://www.amazon.science/research-awards" target="_blank">Amazon Research Award</a>!</p> <p>Paper “Density-Regression: Efficient and Distance-Aware Deep Regressor for Uncertainty Estimation under Distribution Shifts” got accepted in AISTATS 2024.</p> <p>Paper “Addressing the Binning Problem in Calibration Assessment through Scalar Annotation” got accepted in <a href="https://transacl.org/index.php/tacl" target="_blank">Transactions of the Association for Computational Linguistics</a>.</p> <p>Collaboration with <a href="https://suchisaria.jhu.edu/" target="_blank">Suchi Saria</a> received a grant from the <a href="https://www.moore.org/" target="_blank">Gordon and Betty Moore Foundation</a> on safety monitoring of clinical machine learning devices.</p> <p>Paper “Designing for Appropriate Reliance: The role of AI Uncertainty Presentation, Initial User Decisions, and Demographics in AI-Assisted Decision Making” got accepted in <a href="https://cscw.acm.org/2024/" target="_blank">CSCW 2024</a>.</p> <p>Received a grant from <a href="https://iaa.jhu.edu/" target="_blank">JHU IAA</a> to support an AI Fairness Auditing project with <a href="https://tas.ac.uk/" target="_blank">UKRI TAS-Hub</a>.</p> <p>Received the JHU <a href="https://research.jhu.edu/major-initiatives/discovery-awards/2023-awardees/" target="_blank">Discovery Award</a>.</p> <p>I am co-organizing the <a href="https://sites.google.com/view/safe-rl-2023/home?authuser=0" target="_blank">2nd Safe RL workshop</a> in <a href="https://ijcai-23.org/" target="_blank">IJCAI 2023</a>. CFP <a href="https://sites.google.com/view/safe-rl-2023/call-for-papers" target="_blank">here</a>! Please distribute the news and contribute a paper!</p> <p>Paper “Addressing Efficiency Bottlenecks of Conformal Prediction under Standard and Feedback Covariate Shift” got accepted in the <a href="https://icml.cc/Conferences/2023/Dates" target="_blank">ICML2023</a> conference.</p> <p>Paper “Double-Weighting for Covariate Shift Adaptation” got accepted in the <a href="https://icml.cc/Conferences/2023/Dates" target="_blank">ICML2023</a> conference. <a href="https://arxiv.org/abs/2305.08637" target="_blank">Paper</a>.</p> <p>Paper “Learning Calibrated Uncertainties for Domain Shift: A Distributionally Robust Learning Approach” got accepted in the <a href="https://ijcai-23.org/" target="_blank">IJCAI 2023</a> conference.</p> <!-- Paper "Towards human-compatible autonomous car: A study of non-verbal Turing test in automated driving with affective transition modelling" got accepted in [IEEE Transactions on Affective Computing](https://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=5165369). [Paper](https://ieeexplore.ieee.org/abstract/document/10131963). Received the seed grant from JHU [IDIES](https://www.idies.jhu.edu/). [Kelun Lu](https://education.jhu.edu/directory/kelun-lu/) is going to present our work in [2023 Annual Meeting of American Sociological Association](https://www.asanet.org/annual-meeting/2023-annual-meeting/): Learning Processes and Outcomes in An Intelligent Tutoring Platform: A Deep Neural Network Analysis. --> <!--Gave a talk in [Hopkins Population Center](https://popcenter.jhu.edu/). --> <!-- Gave the keynote talk in the JHU [25th Annual Department of Pathology Young Investigators' Day](https://pathology.jhu.edu/research/young-investigators-day). <!--Gave a talk on AI bias and fairness in the JHU WSE [HEEP program](https://engineering.jhu.edu/admissions/graduate-admissions/heep/). --> <!-- Paper "Eyes Are the Windows to AI Reliance: Towards Real-Time Human-AI Reliance Assessment" will be presented at the [Workshop on Trust and Reliance in AI-Assisted Tasks (TRAIT)](https://chi-trait.github.io/#/) at [CHI 2023](https://chi2023.acm.org/). --> <!-- Paper "Distributionally Robust Policy Gradient for Offline Contextual Bandits" got accepted in the [AISTATS 2023](http://aistats.org/aistats2023/) Conference. [Paper](https://proceedings.mlr.press/v206/yang23f.html). Received the [JHU-Amazon AI2AI faculty award](https://ai2ai.engineering.jhu.edu/2022-2023-faculty-research-awards/). Paper "Calibrating Zero-Shot Cross-Lingual (Un-)Structured Predictions" accepted in the [EMNLP 2022](https://2022.emnlp.org/) Conference. [Paper](https://preview.aclanthology.org/emnlp-22-ingestion/2022.emnlp-main.170.pdf). --> <!-- Paper "Ambiguous Images With Human Judgments for Robust Visual Event Classification" accepted in the [NeurIPS 2022 Datasets and Benchmarks Track](https://nips.cc/Conferences/2022/Schedule?type=Poster). Check our [data website](https://katesanders9.github.io/ambiguous-images/) and [paper](https://arxiv.org/abs/2210.03102). Paper "JAWS: Auditing Predictive Uncertainty under Covariate Shift" accepted in the [NeurIPS 2022] (https://nips.cc/Conferences/2022/Schedule?type=Poster) conference. [Paper](https://arxiv.org/abs/2207.10716). --> <!-- Gave a talk at the [RSS2022 Workshop: Risk-aware Decision Making: from Optimal Control to RL](https://sites.google.com/nyu.edu/risk-aware-decision-making). [Paper](https://arxiv.org/abs/2207.12876) accepted in the [ICML2022 Workshop on Spurious Correlations, Invariance, and Stability](https://sites.google.com/view/scis-workshop/home). Gave a talk at the [CVPR2022 UG2+ Challenge Workshop](http://cvpr2022.ug2challenge.org/). Gave a talk at the [ICLR2022 Workshop: Socially Responsible Machine Learning](https://iclrsrml.github.io/). Received [AITC](https://aitc.jhu.edu/) research grant. More details coming. Received [OIDA](https://www.industrydocuments.ucsf.edu/opioids/) research grant. Learn more about the [opioid epidemic](https://en.wikipedia.org/wiki/Opioid_epidemic) and our Opioid Industry Documents Archive (OIDA) [project] (https://hopkinshistoryofmedicine.org/2022/05/11/opioid-industry-documents-archive-featured-in-the-washington-post/). We are working on using machine learning to generate metadata automatically and reliably to help the better usage of the OIDA data in combating the epidemic. --> <div class="row"> </div> </div> </div> </div> </section> <section id="projects" class="home-section"> <div class="container"> <div class="row"> <div class="col-12 col-lg-4 section-heading"> <h2>Featured Projects</h2> </div> <div class="col-12 col-lg-8"> <div class="row isotope projects-container js-layout-row"> <div class="col-lg-12 project-item isotope-item " itemscope itemtype="http://schema.org/CreativeWork"> <i class="far fa-copy pub-icon" aria-hidden="true"></i> <span class="project-title"> <a href="/project/drl/"> Distributionally Robust Learning under Covariate Shift </a> </span> <p class="project-summary">This project covers a series of my work, ranging from fundamentals of distributionally robust learning under covariate shift, to its integration to real-world safe exploration and domain adaption tasks. Media Coverage: <a href="https://engineering.uic.edu/about/coe-news/rise/rise/the-value-of-saying-i-dont-know/" target="_blank">The Value of Saying ‘I Don’t Know’</a>.</p> </div> <div class="col-lg-12 project-item isotope-item " itemscope itemtype="http://schema.org/CreativeWork"> <i class="far fa-copy pub-icon" aria-hidden="true"></i> <span class="project-title"> <a href="/project/social/"> UQ for AI Safety and Fairness </a> </span> <p class="project-summary">We aim to tackle two key challenges in model auditing for safeguarding AI. The first is the ubiquitous distribution shift, especially subpopulation shift. The second is that many UQ approaches require either intensive computing power or an impractical amount or quality of data that may be unavailable in real-world scenarios. Media Coverage: <a href="https://www.cs.jhu.edu/news/putting-trust-to-the-test/" target="_blank">Putting trust to the test</a>.</p> </div> </div> </div> </div> </div> </section> <section id="contact" class="home-section"> <div class="container"> <div class="row"> <div class="col-12 col-lg-4 section-heading"> <h2>Contact</h2> </div> <div class="col-12 col-lg-8"> <ul class="fa-ul" itemscope> <li> <i class="fa-li fas fa-envelope fa-2x" aria-hidden="true"></i> <span id="person-email" itemprop="email"><a href="mailto:aliuATcsDOTjhuDOTedu">aliuATcsDOTjhuDOTedu</a></span> </li> <li> <i class="fa-li fas fa-map-marker fa-2x" aria-hidden="true"></i> <span id="person-address" itemprop="address">Malone Hall, 3400 N. Charles Street, Baltimore, Maryland 21218</span> </li> </ul> </div> </div> </div> </section> <div class="container"> <footer class="site-footer"> <p class="powered-by"> © 2022 · Anqi (Angie) Liu <span class="float-right" aria-hidden="true"> <a href="#" id="back_to_top"> <span class="button_icon"> <i class="fas fa-chevron-up fa-2x"></i> </span> </a> </span> </p> </footer> </div> <div id="modal" class="modal fade" role="dialog"> <div class="modal-dialog"> <div class="modal-content"> <div class="modal-header"> <h5 class="modal-title"></h5> <button type="button" class="close" data-dismiss="modal" aria-label="Close"> <span aria-hidden="true">×</span> </button> </div> <div class="modal-body"> <pre><code class="tex hljs"></code></pre> </div> <div class="modal-footer"> <a class="btn btn-outline-primary my-1 js-copy-cite" href="#" target="_blank"> <i class="fas fa-copy"></i> </a> <a class="btn btn-outline-primary my-1 js-download-cite" href="#" target="_blank"> <i class="fas fa-download"></i> </a> <div id="modal-error"></div> </div> </div> </div> </div> <script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.3.1/jquery.min.js" integrity="sha512-+NqPlbbtM1QqiK8ZAo4Yrj2c4lNQoGv8P79DPtKzj++l5jnN39rHA/xsqn8zE9l0uSoxaCdrOgFs6yjyfbBxSg==" crossorigin="anonymous"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/jquery.imagesloaded/4.1.3/imagesloaded.pkgd.min.js" integrity="sha512-umsR78NN0D23AzgoZ11K7raBD+R6hqKojyBZs1w8WvYlsI+QuKRGBx3LFCwhatzBunCjDuJpDHwxD13sLMbpRA==" crossorigin="anonymous"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/twitter-bootstrap/4.1.3/js/bootstrap.min.js" integrity="sha256-VsEqElsCHSGmnmHXGQzvoWjWwoznFSZc6hs7ARLRacQ=" crossorigin="anonymous"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/jquery.isotope/3.0.4/isotope.pkgd.min.js" integrity="sha512-VDBOIlDbuC4VWxGJNmuFRQ0Li0SKkDpmGyuhAG5LTDLd/dJ/S0WMVxriR2Y+CyPL5gzjpN4f/6iqWVBJlht0tQ==" crossorigin="anonymous"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/fancybox/3.2.5/jquery.fancybox.min.js" integrity="sha256-X5PoE3KU5l+JcX+w09p/wHl9AzK333C4hJ2I9S5mD4M=" crossorigin="anonymous"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/9.12.0/highlight.min.js" integrity="sha256-/BfiIkHlHoVihZdc6TFuj7MmJ0TWcWsMXkeDFwhi0zw=" crossorigin="anonymous"></script> <script src="/js/hugo-academic.js"></script> <script>hljs.initHighlightingOnLoad();</script> <script> const search_index_filename = "/index.json"; const i18n = { 'placeholder': "", 'results': "", 'no_results': "" }; const content_type = { 'post': "", 'project': "", 'publication' : "", 'talk' : "" }; </script> <script id="search-hit-fuse-template" type="text/x-template"> <div class="search-hit" id="summary-{{key}}"> <div class="search-hit-content"> <div class="search-hit-name"> <a href="{{relpermalink}}">{{title}}</a> <div class="article-metadata search-hit-type">{{type}}</div> <p class="search-hit-description">{{snippet}}</p> </div> </div> </div> </script> <script src="https://cdnjs.cloudflare.com/ajax/libs/fuse.js/3.2.1/fuse.min.js" integrity="sha256-VzgmKYmhsGNNN4Ph1kMW+BjoYJM2jV5i4IlFoeZA9XI=" crossorigin="anonymous"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/mark.js/8.11.1/jquery.mark.min.js" integrity="sha256-4HLtjeVgH0eIB3aZ9mLYF6E8oU5chNdjU6p6rrXpl9U=" crossorigin="anonymous"></script> <script src="/js/search.js"></script> </body> </html>