Dan Spencer doesn't know Quarto yet

102 posts

Dan Spencer doesn't know Quarto yet

Dan Spencer doesn't know Quarto yet

@danCMDstat

"Perfection is achieved, not when there is nothing more to add, but when you put comments in your code." ~Antoine de Saint-Exupery, today

Katılım Eylül 2020
65 Takip Edilen53 Takipçiler
Samuel Davenport
Samuel Davenport@BrainStatsSam·
@danCMDstat To generate more realistic noise simulations I recommend bootstrapping resting state with fake task a la Eklund 2016.
English
1
0
1
0
Dan Spencer doesn't know Quarto yet retweetledi
Mathieu
Mathieu@miniapeur·
Mathieu tweet media
ZXX
8
340
2.8K
0
Samuel Davenport
Samuel Davenport@BrainStatsSam·
@hufengling @danCMDstat Yeah I wonder how realistic the images are that were generated in practice. Difficult to really know. It seems safer to me to stick to validations using real data where we know that the distribution of the noise/signal is realistic.
English
1
0
0
0
Dan Spencer doesn't know Quarto yet
This is a truly genius idea. As a statistician working in neuroimaging methods, there are lots of times when you want simulated data that are close to the real thing, but with a controllable ground truth so you can test model power and accuracy. Great job!
Walter Hugo Lopez Pinaya 🍍@Warvito

🎉New preprint! We used Diffusion Models (same one from #dalle2 and @StableDiffusion ) to generate 3D MRI images of the brain conditioned on several covariates and make 100k synthetic brains openly available Paper: arxiv.org/abs/2209.07162 Dataset: tinyurl.com/32p4hu7d 1/n

English
1
0
4
0
Dan Spencer doesn't know Quarto yet
@BrainStatsSam Yes, but then you are limited to the data that have been collected. I agree that you *can* and many people *have* simulated using the UK Biobank data. I think this is likely the next step. You have no limit to the number of subjects, plus you can simulate over unseen conditions.
English
1
0
1
0
Samuel Davenport
Samuel Davenport@BrainStatsSam·
@danCMDstat I think a better approach to generating realistic simulations is to leverage big datasets like the UK biobank (i.e. setting aside a large number of subjects to calculate a ground truth and dividing the remaining subjects into small groups on which to test the methods).
English
2
0
2
0
Ray Bai
Ray Bai@raybai07·
The 2022-2023 academic job market is upon us! Resharing my post from Feb. '20 about my experience on the TT academic job market. Includes info on putting together job application materials, preparing for virtual + on-site interviews and job talk. (1/2) raybai.net/my-experience-…
English
3
0
9
0