addressalign-toparrow-leftarrow-rightbackbellblockcalendarcameraccwcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-checkcircle-with-crosscircle-with-pluscrossdots-three-verticaleditemptyheartexporteye-with-lineeyefacebookfolderfullheartglobegmailgooglegroupsimageimagesinstagramlinklocation-pinm-swarmSearchmailmessagesminusmoremuplabelShape 3 + Rectangle 1outlookpersonJoin Group on CardStartprice-ribbonShapeShapeShapeImported LayersImported LayersImported Layersshieldstartickettrashtriangle-downtriangle-uptwitteruserwarningyahoo

Fast Quantification of Uncertainty and Robustness with Variational Bayes

  • This location is shown only to members

  • Our December meeting will feature a talk from Tamara Broderick, Assistant Professor at MIT, about Variational Bayes techniques. We'll also make 10 mins for lightning talks if anyone wants to get up and quickly tell us all about anything cool they're working on / new packages etc. (Let us know before the meeting). 

    This is a joint event with our friends of the Boston Algorithmic Trading group. The event is sponsored by Quantopian and Microsoft.


    Fast Quantification of Uncertainty and Robustness with Variational Bayes


    In Bayesian analysis, the posterior follows from the data and a choice of a prior and a likelihood. These choices may be somewhat subjective and reasonably vary over some range. Thus, we wish to measure the sensitivity of posterior estimates to variation in these choices. While the field of robust Bayes has been formed to address this problem, its tools are not commonly used in practice---at least in part due to the difficulty of calculating robustness measures from MCMC draws. We demonstrate that, by contrast to MCMC, variational Bayes (VB) techniques are readily amenable to fast robustness analysis. Since VB casts posterior inference as an optimization problem, its methodology is built on the ability to calculate derivatives of posterior quantities with respect to model parameters. We use this insight to develop local prior robustness measures for mean-field variational Bayes (MFVB), a particularly popular form of VB due to its fast runtime on large data sets. A potential problem with MFVB is that it has a well-known major failing: it can severely underestimate uncertainty and provides no information about covariance. We generalize linear response methods from statistical physics to deliver accurate uncertainty estimates for MFVB---both for individual variables and coherently across variables. We call our method linear response variational Bayes (LRVB).

    Speaker Bio:

    Tamara Broderick is the ITT Career Development Assistant Professor in the Department of Electrical Engineering and Computer Science at MIT. She is a member of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Institute for Data, Systems, and Society (IDSS). She completed her Ph.D. in Statistics with Professor Michael I. Jordan at the University of California, Berkeley in 2014. Previously, she received an AB in Mathematics from Princeton University (2007), a Master of Advanced Study for completion of Part III of the Mathematical Tripos from the University of Cambridge (2008), an MPhil by research in Physics from the University of Cambridge (2009), and an MS in Computer Science from the University of California, Berkeley (2013). Her recent research has focused on developing and analyzing models for scalable, unsupervised machine learning using Bayesian nonparametrics. She has been awarded the Evelyn Fix Memorial Medal and Citation (for the Ph.D. student on the Berkeley campus showing the greatest promise in statistical research), the Berkeley Fellowship, an NSF Graduate Research Fellowship, a Marshall Scholarship, and the Phi Beta Kappa Prize (for the graduating Princeton senior with the highest academic average).


    Twitter: @BostonBayesians
    Email: [masked]


    • 6:00pm: Networking. Food and drink sponsored by Quantopian  
    • 6:30pm: Talk by Tamara Broderick and Q&A
    • 7:30pm: Networking
    • 8:30 End of event.

    Note to all attendees: 

    All attendees must RSVP before the event. When arriving at the event, all attendees must register with building security, located in the lobby of One Memorial Drive, prior to entering Microsoft’s facility. Please present a government issued, photo ID to lobby security and sign a waiver in order to be allowed access to the building. 

    A note from our sponsor Quantopian:

    Quantopian inspires talented people from around the world to write investment algorithms. Quantopian provides capital, data, and infrastructure to algorithm authors. Quantopian offers license agreements for algorithms that fit its investment strategy, and the licensing authors are paid based on their strategy’s individual performance. Quantopian provides everything a quant needs to create a strategy and profit from it. Learn more, visit us at:

    A note from our sponsor Microsoft:

    Microsoft New England (MSNE) is a major center for technical innovation and research. Located in the heart of Cambridge, Massachusetts, the MSNE campus includes two buildings—One Memorial Drive and the recently renovated One Cambridge Center. At the core of MSNE is the Microsoft New England Research & Development Center (NERD). NERD is a world-renowned research and software development center, and is home to teams working on critical products and services like Microsoft Office 365 and Microsoft SQL. NERD also serves as a center of gravity for the local tech community, having hosted more than 1,000 events and welcomed more than 100,000 visitors since opening in 2008. 

Join or login to comment.

Want to go?

Join us!

112 going

28 spots available

Our Sponsors

People in this
Meetup are also in:

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy