Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Denis A. Engemann Profile picture
Denis A. Engemann
@dngman
Neuroscientist in Pharma R&D @Roche. Passionate about electrophysiology (M/EEG), cognition, machine learning, stats, (...). Tweets = personal views & opinions.
Dec 25, 2020 • 9 tweets • 6 min read
Encouraging! #AppleM1 Silicon (MBA) smokes my 2017 MBP15" i7 on #rstats #tidverse tidymodels hotel example, random forests (last fit 100 trees). Experimental arm-R build = extra speedup. Thanks @fxcoudert for gfortran build & @juliasilge @topepos + team for the nice API + DOC. And it’s wonderful to see that essential R packages are working on the M1 platform.
May 19, 2020 • 10 tweets • 5 min read
I am very excited to share our latest work published in @eLife! We combined #MEG, #fMRI and #MRI for #BrainAge prediction & #biomarker development. Each modality added unique information & enhanced brain-behavior mapping! elifesciences.org/articles/54055 Thread👇 1/ We combined anatomical MRI (surface area, thickness, volume), fMRI-connectivity and MEG (source power, connectivity, alpha-peak, 1/f, latencies) with a stacking approach (1: ridge, 2: random forest). We made sure our model potentially extracts information from missing values.
May 18, 2020 • 12 tweets • 5 min read
Excited to share our paper @NeuroImage_EiC by @DavSabbagh with @PierreAblin @GaelVaroquaux @agramfort doi.org/10.1016/j.neur… Nonlinear subject-level regression on M/EEG using linear models without source localization: theory + empirical benchmarks Thread👇 1/ When regressing outcomes on M/EEG power spectra (f = power; f = log(power), ...), volume conduction creates a nonlinear problem that cannot be addressed with otherwise effective linear models. Source localization can fix this but is not always available. How can we do without?