@dtfraz

Statistician/Econometrician working on simulation-based inference at @MonashUni (EBS).

 Melbourne, Victoria
dtfrazier.netlify.com
Joined November 2019
  • Tweets213
  • Following112
  • Followers520
8Media
David T. Frazier retweeted
Chris Drovandi@chris_drovandi10 Jun 2023
Our paper on improving marginal approximations in likelihood-free inference now online at JCGS. @dtfraz @QUTDataScience tandfonline.com/doi/full/10.…

Improving the Accuracy of Marginal Approximations in Likelihood-Free Inference via Localization

Likelihood-free methods are an essential tool for performing inference for implicit models which can be simulated from, but for which the corresponding likelihood is intractable. However, common li...

tandfonline.com
6
25
2,700
I just wanted to thank everyone for all the well-wishes I've received over the last week. I especially want to thank @Science_Academy for awarding me the Moran medal, and @robjhyndman for nominating me. Also, a big congratulations to my co-winner Rachel Wang. Thanks again!
6
4
32
3,739
Who's interested in a cautionary tale of when you should and shouldn't use VB methods in SSMs? If that's the kinda thing you're into, check out our recent paper in #JCGS with @RubenLoaizaMay1 and Gael Martin: tandfonline.com/doi/abs/10.1…

Variational Bayes in State Space Models: Inferential and Predictive Accuracy

Using theoretical and numerical results, we document the accuracy of commonly applied variational Bayes methods across a range of state space models. The results demonstrate that, in terms of accur...

tandfonline.com
1
5
18
In short, if your goal is inference, then you probably shouldn't use fixed form VB approximations in SSMs. Even very good ones are inconsistent. However, if your goal is prediction, and you don't have a long time series, then VB is surprisingly accurate.
1
5
Amsterdam is absolutely lovely, how have I never been here before?!?!?
1
20
Big thanks to @chris_drovandi for leading this "eye-opening" work on the use of marginal approximations in ABC!
Chris Drovandi@chris_drovandi17 Jul 2022
Our new pre-print with @dtfraz and David Nott examines marginal ABC approximations based on low-dimensional summaries 🧵👇. @QUTDataScience arxiv.org/abs/2207.06655

Improving the Accuracy of Marginal Approximations in...

Likelihood-free methods are an essential tool for performing inference for implicit models which can be simulated from, but for which the corresponding likelihood is intractable. However, common...

arxiv.org
7
Very excited for Pierre Jacob's "Susie Bayarri Lecture" on modular inference at #ISBA2022. His work has touched on so many different aspects of Bayesian statistics it is astounding! Truly a fantastic scholar!
22
Great talk at #ISBA2022 by Prof. Mike West on the dominance of Bayesian Predictive Synthesis in forecasting: moving Bayesian prediction beyond "standard" approaches like BMA, etc. It was a surprise and honour to have Mike acknowledge our work on "Focused Bayesian prediction."
5
3
24
The BPS approach is definitely the way to go when you have specific forecasting goals in mind, but want to be Bayesian in your decision making/thinking. But don't take my work for it, check out their impressive results.
2
Nice talk on posterior concentration of ABC using integral probability metrics by @SirioLegramantiv #ISBA2022. Seems promising!! @PierreAlquier
3
2
1
31
Looking forward to reading this and picking @PierreAlquier brain afterwards! Nice to see some more theory on this aspect of ABC.
This Post is from an account that no longer exists. Learn more
5
Interested in whether to use likelihood-free methods based on summary statistics, or methods based on empirical probability measures? Check out this great work led by @chris_drovandi for helpful ideas on when to use one method or the other.
Chris Drovandi@chris_drovandi21 May 2022
Excited that our paper on comparing likelihood-free methods with and without summary statistics is now available Open Access with Statistics and Computing. Enjoyed working with @dtfraz on this. @QUTDataScience link.springer.com/article/10…

A comparison of likelihood-free methods with and without summary statistics

Statistics and Computing - Likelihood-free methods are useful for parameter estimation of complex models with intractable likelihood functions for which it is easy to simulate data. Such models are...

link.springer.com
5
Delighted to announce that my joint work with @chris_drovandi, @robertjk59 and David Nott on the asymptotic behavior of Bayesian Synthetic Likelihood is now published at JASA! Big thanks to @chris_drovandi for making this open access for all to enjoy! tandfonline.com/doi/abs/10.1…

Bayesian Inference Using Synthetic Likelihood: Asymptotics and Adjustments

Implementing Bayesian inference is often computationally challenging in complex models, especially when calculating the likelihood is difficult. Synthetic likelihood is one approach for carrying ou...

tandfonline.com
7
32
Many thanks to @JacobPriddle for leading this empirically useful and interesting project! Also, big thanks to @chris_drovandi and @ABC_Research for their hard work on the project!
Jacob Priddle@JacobPriddle26 Oct 2021
Very excited that the final proofs of our wBSL paper are finally online! A huge thanks goes to my amazing co-authors for making this work possible! @chris_drovandi @ABC_Research @dtfraz And here is my link for (another) 50 free copies: tandfonline.com/eprint/ZAGFA…

Efficient Bayesian Synthetic Likelihood With Whitening Transformations

Likelihood-free methods are an established approach for performing approximate Bayesian inference for models with intractable likelihood functions. However, they can be computationally demanding. B...

tandfonline.com
6
Wow! This looks great! If your interested in learning about PAC-Bayes, check this out. Thanks to @PierreAlquier for doing the hard work!
This Post is from an account that no longer exists. Learn more
2
Very cool award. It's great to see this side of statistics getting the recognition it deserves.
SSA Victoria and Tasmania@SSAVicTas27 Sep 2021
📢Announcing the Di Cook award by SSA Vic: This is an award acknowledging excellent open-source statistical software developed by students (or recent graduates) of Victorian and Tasmanian institutes. Submissions close on 26th Nov. See statsocaus.github.io/dicook-… for guidelines.

Di Cook Award: SSA SCV Di Cook Award

Open-Source Statistical Software Award by Section on Statistical Computing and Visualisation

statsocaus.github.io
9
Holy crap! Just had an earthquake in Seaford Vic! My first one! #seaford #Melbourne
3
2
BIG Congratulations to @chris_drovandi!! Very well-deserved!!
Australian Research Council
@arc_gov_au10 Aug 2021
#ARCNetworkMsg: Announcement of grant outcomes under the #FutureFellowships 2021 #FT21 scheme ow.ly/h3gR50FNN3M #ARCAnnouncementKit also available ow.ly/ZrLI50FNN3N For funded projects please see ow.ly/G93y50FNN3O
1
15
David T. Frazier retweeted
Matias Quiroz@MQBayes20 Jul 2021
Very excited (and honoured!) to teach a course in Subsampling MCMC at @DiscoverAMSI's Winter School. A lot of preparation has come to an end and it is finally happening in about three hours!
1
4
32
Congratulations to @robjhyndman for being awarded the prestigiouse Pittman medal from @StatSocAus!! statsoc.org.au/News-and-medi… @ACEMathStats @MonashEBS
1
1
2
40
Load more