Early fits did not establish lower loss than full rest correction with a 10-trial half-life in eight SEEG readers
We tested if early labels could set a useful strength for a past-rest correction in eight fixed SEEG readers. The data came from public OpenNeuro ds004473, version 1.0.2. The primary rest update had a half-life of 10 retained trials. Early fits chose zero correction in five readers. On 975 later trials after task condition changes, fitted minus full loss was -0.0147 bits per window. The descriptive 95 percent person-bootstrap interval was [-0.0327, +0.0041]. The fitted correction did not improve on the fixed reader. A five-trial sensitivity check reduced loss relative to full correction, but did not improve on the fixed reader. These exploratory results use a prior cohort and response-timed windows. The package preserves features, model states, predictions, four figures, code, source hashes, and reproduction instructions.
stereoelectroencephalography · SEEG · brain-computer interface · logistic regression · calibration · distribution shift · log loss · OpenNeuro
References
5- usesA ten-trial rest update did not establish higher later-window SEEG accuracy in eight patients at thresholds set for 10 percent rest false positives sci:sha256:a2ab4f25dd9e0d622d6cfafdf4e6924fd0589a02a21912150ae1a6c813a707b0
Its ten-trial rest update did not establish a mean later accuracy gain at rest-based thresholds. This result motivated testing whether early labels support full correction. The present fit, split, threshold, and past-input timing differ.
- usesEight fixed SEEG decoders scored 65.8% mean balanced accuracy for response versus rest windows after task condition changes sci:sha256:27c3a162d5f23c03418feb1337bb50c3dbd6fcf78da3fdc3bdc12da973491084
This publication supplies all eight feature sets, primary fixed reader states, trial-fit indices, and source metadata. We reuse these files with hash checks. We add an early bounded correction-strength fit and test the next retained condition.
- usesA 32-trial rest covariance update did not establish higher SEEG balanced accuracy than a mean update in eight people sci:sha256:b3fe71ef5b07a9e3b47dfadeb06c62168bfaa85e7e4bc609b917c407f8abe4eb
Its comparison of rest mean and covariance updates did not establish a mean later accuracy gain. It motivated a bounded-strength check. It uses another contact panel and split, so it is not an independent replication of this result.
- usesdoi:10.18112/openneuro.ds004473.v1.0.2
The analyzed source is OpenNeuro ds004473 version 1.0.2; its exact-version metadata records this DOI and a CC0 license.
- usesdoi:10.1088/1741-2552%2Facae0a
This paper describes the source SEEG recordings and forced two-choice task used in the analysis.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.