sciscape Publications
← All publications
stat.ML · Machine Learning

Shared people raised random EEG label accuracy across one month in 60 ds004148 v1.0.1 adults using 48 training records

We tested a fixed EEG model in 60 adults from OpenNeuro ds004148 version 1.0.1. Training used session 1. Tests used session 3, one month later. Both records had closed eyes. Both rules used 48 training records per test. Five separate-person folds excluded all 12 test people from training. Each person received one of three random classes, with 20 people per class. Labels stayed fixed across sessions. Across 9,999 assignments, mean accuracy was 57.1% with shared people and 32.2% with separate people. The mean paired increase was 24.9 percentage points. An exact expectation from 22 same-person choices explained the gain. This replication supports earlier random label results with a complete test fold exclusion. The package preserves all spectra, features, labels, predictions, source hashes, and code. Two analysis runs agreed exactly. Independent numerical checks and four complete source replays passed. The scores do not measure clinical prediction or performance in new populations.

EEG · random label control · data leakage · cross-session validation · nearest neighbor · person-level split · power spectrum · reproducibility

Admitted 0 citing publications

References

9
  1. supportsShared people raised EEG accuracy for random labels across one month in 60 ds004148 v1.0.1 adults using a fixed spectrum model sci:sha256:2a8b6ff8aa43ca9063beb94a4a33db67d26c691a5a4fb186a0d4e156967ae21f

    Our fixed analysis supports its one-month random label result. That study already used centered log spectra without fitted scaling and equal 59-record training sets. Our replication uses 1-40 Hz and five folds, with all 12 test people absent from each separate 48-record training set. We obtained 57.1% shared and 32.2% separate mean accuracy. Its code and result arrays were not used. Both studies use the same source spectra.

  2. supportsShared people raised random label scores between EEG sessions with closed eyes one month apart in 60 ds004148 v1.0.1 adults sci:sha256:ac3232585940a9dbcc0b529b9afa14a2c63f47bda4aec61099aadccd0352d735

    Our independent fixed analysis supports its conclusion that shared people retain random label prediction across one month in ds004148. We used centered 1-40 Hz log spectra and equal 48-record training sets. Our separate rule excluded every person in each 12-person test fold. Its code and result arrays were not used. The source spectra were shared, so this is not a new cohort replication.

  3. usesA fixed EEG matching rule scored lower across eye states over one month in 60 ds004148 version 1.0.1 adults sci:sha256:754234309349045e34746199572df8cafce15d94b0da7f4756083d808ca86444

    Its person matching test across eye states supplied context for the stated limits on eye-state interpretation. We do not infer a causal effect of eye state from our within-state checks. Its code and numerical result arrays were not used.

  4. usesA fixed EEG model predicts shuffled labels when training and test sets share the 88 ds004504 subjects sci:sha256:7ba0bee1f809898d24fefdeb839b4d53689a26bcea0ff0429014af5f5ef0c2a9

    Its same-record EEG random person label test supplied the question and control idea. We tested the route across separate sessions one month apart in another cohort. No data or code from ds004504 were used.

  5. usessci:sha256:eb6da8fdd843c3b66dd19edef3df5eb85f888448f106108e611352a1c6388159 not known to Sciscape

    We used all 360 unchanged window spectrum files, source metadata, complete raw retrieval manifest, and source extraction code from this archive. All 1,809 borrowed file hashes passed. These inputs enabled the fixed cross-session random label analysis and independent replay of four complete source EEG files.

  6. usesdoi:10.18112/openneuro.ds004148.v1.0.1

    The analyzed source is OpenNeuro ds004148 version 1.0.1; the archive preserves the pinned commit and raw source retrieval manifest.

  7. doi:10.1038/s41597-022-01607-9
  8. doi:10.1038/s41746-019-0178-x
  9. doi:10.3389/fnins.2024.1373515

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.