Nominations for 2023 Stan Governing Body

As announced in Elections for the next Stan Governing Body , we will use the Stan forums for nominations for the 2023 Stan Governing Body (SGB).

Please respond to this post to self-nominate for either a 1-year or 2-year term on SGB before 2022-11-15T08:00:00Z. We encourage all nominees to briefly summarize their experiences with Stan and their goals for the SGB. Feel free to add links to any content you think is relevant. And if you know someone who you think would be a good fit for the SGB please let them know and encourage them to nominate themselves.

For other details about the election and links to information about the SGB see Elections for the next Stan Governing Body . For examples of past nominations, here is last year’s thread.


Yi Zhang on behalf of @SGB

4 Likes

On behalf of @SGB I have this nomination thread open for another two weeks. After much deliberation the SGB has made the decision that starting from this election the governing body will be of a flexible size between 3 to 5 each year.

5 Likes

Dear all,
I’m Charles Margossian, and I’d like to run for a 1 year term for the SGB.

EDIT: I am also interested in running for 2 year term.

Put plainly: I owe the Stan community an immense debt of gratitude. My interactions with developers, users, and researchers within and around the project have not only introduced me to Bayesian statistics but further pushed me to pursue research in this field. Stan is an empowering tool, helping individuals learn principled statistics and make the most of their data. We are connecting researchers, developers, and modelers across a broad palette of fields. This is the mission we need to uphold.

My general sense is that there is a lot of great work happening within the Stan community, at all levels: (i) applications, (ii) software development and (iii) methodological research. I would like to facilitate conversations between actors in all these areas. To that end, the actions I propose to pursue are the following:

  • Resume StanCon. Find ways to make the conference more accessible, for example with hybrid online options and by picking a diversity of locations.
  • Continue the StanConnect events.
  • Support members of the community to give workshops on Stan – either at special events, conferences, or schools.

Crucially such events are an opportunity to uphold the SGB’s commitment to diversity, by fostering a safe and inclusive environment, giving representation to otherwise underrepresented groups, and extending accessibility to the Stan project.

Experience

I have been a Stan developer for 7 years. My contributions have mostly been geared towards supporting implicit functions in Stan, such as ODE solvers, algebraic solvers, and hidden Markov models. I’m also the co-developer of Torsten, an extension of Stan for pharmacometrics.

Currently, I’m a postdoc at the Flatiron Institute, Center for Computational Mathematics. I recently completed my PhD in statistics at Columbia University. My research covers topics, such as MCMC, variational inference, integrated Laplace approximations and automatic differentiation. I’ve also been able to collaborate with practitioners in various fields, notably Stan users in pharmacometrics and epidemiology.

Depending on the need of the community, I’m open to running for a 2 year term.

Thank you for your consideration,
Charles

15 Likes

Summary

I passionately believe that Stan has already proven to be a ground-breaking open source project that has unlocked a whole raft of practical applications for Bayesian inference. However, I think Stan is currently only scratching the surface of its potential to make a difference. To continue to grow, I believe Stan needs to embrace a desire to continually evolve. I would like to be involved in the SGB as I believe that my experience can help unlock Stan true potential.

Vision

I’d like to serve on the SGB to enable Stan to have broader applicability, utility and use.

NUTS was a revolutionary advance. By making it possible to apply NUTS to a diverse range of applications, Stan has made it possible for a vast range of users to use Bayesian models to extract information they see as important from their data. While the ecosystem around Stan has served users well, because of Stan’s tight-coupling to problems that can be tackled with NUTS, Stan’s applicability is limited to problems that can be posed as being dependent on fixed-dimensionality vector-valued parameters.

Other problems (eg related to dynamic problems and discrete parameters) exist. Not only this, but users are identifying that they need to solve these problems. Meanwhile, algorithms are being proposed as solutions. My personal view is that the Stan ecosystem makes it hard for these people (both the users posing these problems and the algorithm developers proposing solutions) to fully exploit Stan’s potential: users can misinterpret guidance to imply that they are posing the wrong problem; in my experience at least, algorithm developers can struggle to contribute in the presence of the combination of well-rehearsed arguments to maintain the algorithmic status-quo and a seemingly-impenetrable codebase (which makes it hard to accumulate empirical evidence to counter those arguments).

It is clear to me that the Stan developer community does not want this situation to exist or to persist. Indeed, I am very optimistic that there is hope! More specifically, I am enthused by the recent development of BridgeStan [BridgeStan 1.0.0 released]. BridgeStan provides algorithm developers with the ability to try out their approaches in the context of Stan models. I believe BridgeStan will ease the process of objectively comparing approaches and help provide an environment where users can be provided with solutions to problems that they cannot currently tackle with Stan. Put simply: BridgeStan is an important step to enabling users to tackle problems that Stan cannot currently handle.

I am also aware that there are a large number of researchers around the world developing algorithms that could help Stan’s users but which currently create siloed github repos, each of which have somewhat modest user bases. Such researchers do not view themselves as Stan developers but do represent a latent potential to augment the funding and people that are aligned with Stan’s development. If we can align the output of such ongoing research with Stan’s development goals, there is then an increased chance that we can create a virtuous circle whereby we secure funding explicitly aligned with Stan’s development and such funded development achieves more `bang for its buck’.

Streamlining the process of broadening Stan’s applicability, utility and use will require further changes to how Stan is governed, how development is funded and how the community feels. The SGB has a vital role to play in facilitating this shift and I would love to have the opportunity to help make this happen.

Experience

My initial exposure to Stan was as a user. For example, a Stan model that two of my PhD students developed has been used since February 2021 by UK government as one of approaches that contributes to the official calculation of the UK’s R-number for COVID-19. While I have also helped with the organisation of StanCon2020, my primary involvement with Stan has been as the lead for a research group of approximately 60 researhers (PhD students, post-docs and data scientists) at the University of Liverpool in the UK. I have secured funds that have enabled my team to work over recent years, somewhat behind the scenes, to develop software and algorithms pertinent to Stan. These algorithms have been developed to address issues that the potential users I interact with see as gaps in Stan’s applicability. While our contributions to Stan releases have not been extensive (and this is arguably my point), we have recently developed a post-processing tool (currently using BridgeStan) that uses control variates to dramatically reduce the standard error associated with parameter estimates output by Stan at negligible computational cost [Control Variates for Reduction of Sample Variance in Stan]. However, our focus has been on developing Sequential Monte Carlo (SMC) methods that can extend Stan’s applicability in three ways:

  1. users want to exploit available compute (eg GPU farms purchased for Deep Learning) to apply Stan at scale such that they can be significantly more ambitious in terms of the models that they can apply to their data: we are using SMC samplers as an inherently parallel alternative to MCMC to achieve this;
  2. users want to tackle streaming problems [StanCon 2020. Talk 5: Simon Maskell. Towards an Interface for Streaming Stan - YouTube] and calibrate the models used in such contexts: we have adapted Stan to use Fixed-Lag SMC (ie a high performance particle filter) and associated calibration techniques (eg SMC^2);
  3. users want to tackle problems involving model selection (including problems, eg involving decision forests, with unknown dimensionality that could be tackled using Reversible Jump): we are developing a software framework that ensures self-consistency when modelling discrete variables (akin to auto-diff ensuring that the gradient is actually equal to the gradient of a specified function) as well as implementing SMC samplers (and other efficient algorithms, eg HINTS: [HINTS: An alternative to NUTS that doesn't require gradients]) in this context.

I believe this experience means that I am well placed to serve on the SGB to facilitate the vision described above. Given the magnitude of the changes that I view as necessary, I propose to serve for the maximum two year term.

12 Likes

This is a pretty huge change to the governance and I haven’t seen any public discussion about the topic at all. Can the @SGB make a public statement about the decision and the updated structure of the governance?

2 Likes

Dear Community,

my name is Caetano Souto Maior, and I would like to nominate myself for a one-year term in the Stan Governing Body.

I am currently a Postdoctoral Fellow in the Applied Statistics group at the Basque Center for Applied Mathematics (BCAM). My background, however, is mostly in the natural sciences, where I have been not only defending but also applying bayesian inference for almost a decade.
I am faculty at the FAES Graduate School @ NIH where I have taught Linear Algebra with Applications in Statistics, Single- and Multivariate Calculus, and recently developed the Mathematical Modeling for Life Scientists course. I am also a certified trainer at The Carpentries and a volunteer translator for Raspberry Pi.

I nominated myself last year, and although I wasn’t elected to one of four spots, I got an encouraging enough number of votes that I’d do it again this year. Then my proposals were centered around (i) a functional, education-oriented implementation of HMC methods, and (ii) dissemination of Stan in mechanistic modeling in the natural sciences. Although I don’t want to abandon those ideas completely, this time around I want to keep them in the back-burner as part of a broader perspective.

My intended contribution to the Stan Community can be summed up under two main topics:

  1. Better statistical literacy through Bayesian inference: Statistics is often seen by users as an impenetrable labyrinth of arcane rules and unclear use cases. For the initiated, Bayesian Statistics in particular is perceived by some to be a sort of sect requiring deep commitment to subjective philosophical principles or ideologies. Nothing could be farther form reality; if Statistics becomes “easy” with time it’s because experience teaches us that it all essentially boils down to parameter inference under some model(s), and Bayesian inference is especially well suited to paint that general picture. Stan and its community have played an important role in popularizing Bayesian practice and enabling state-of-the-art methods to be adopted widely. My proposal to this point is to improve/consolidate (existing efforts), and expand Stan’s role as a tool for teaching and training to improve statistical literacy broadly;

  2. End the “Bayesian wars” in scientific research: I got into an argument with a self-proclaimed non-frequentist who was critical of what they called “Bayesian propaganda” at a conference this summer. Despite some disagreement on my side I ended up accepting that almost everything related to inference can be done using either a frequentist or bayesian approach (as discussed recently here, MLE can be seen as a particular case of Bayesian inference, for instance). In turn, they conceded that Bayesian practices were usually more solid (possibly because they required a better understanding of the methods). Based on this and other discussions over the years I believe there is an opportunity to promote Bayesian inference (and by extension HMC and Stan as a state-of-the-art implementation) to scientists as simply Statistics and Inference, and do away with the nonsense surrounding “Bayesianism”. Concretely, my proposal here is to promote Bayesian approaches that “mimic” Frequentist methods (something brms, for instance has already done very successfully for linear models) rather than trying to (completely) replace them, avoiding conflict and making the introduction of the more solid practices smoother.

As I see it, these proposals are essentially orthogonal to those of this (and also probably last) year’s nominees, but as before it is also in line with my own and other’s efforts to promote Stan to different audiences in multiple research fields, different languages, in the Global South, and increase the diversity of the community, to mention some of the previous proposals by members of the SGB.

I am available to answer and questions or concerns, as well as to elaborate on the proposals summarized in the points above.

best,
Caetano

9 Likes

The ideal size of the SGB is 5 members, but in the event that there are not enough candidates the SGB will consist of 3 members.

In December, faced with no candidates, the SGB considered how best to go forward. Another concern was that in the face of community apathy, whoever stepped forward could win by default. The decision was made to remain flexible while extending the nomination period.

1 Like

The (extended) nomination period has closed.
Voting will be open Jan 16-Jan 31 (same as 2022).
Transition between the governing bodies will be February 2022.