sciscape Publications
← All publications
cs.LG · Machine Learning

Exact bounds saved 17.7% of pair calculations and kept all full reader decisions in eight SEEG recordings with clipped pair scores

Exact score bounds reduced pair calculations for fixed SEEG task-state readers in eight public recordings. Bounds clipped each pair score to its early-training range. An exact subset program chose a fixed pair order from training windows only. On 1,348 later windows, the method used 6.583 of eight pairs on average. It saved 17.71% of pair calculations, with a descriptive 95% person interval of 15.42% to 20.03%. All stopped decisions matched the full clipped reader. Clipping changed two decisions from the original reader. The exact order gave almost the same later cost as a simple width order. In nine local timing repeats, early processing took 3.24 times the full batch time and 0.88 times the full serial time. Fewer calculations therefore did not show faster batch processing. The package preserves raw selected buffers, fixed models, proofs, exact results, code, three figures, source hashes, and retrieval instructions. Independent rational checks passed, and a new process reproduced all 21 numerical files exactly. These reused, response-aligned task windows do not establish online control, device energy savings, or stimulation benefit. This revision corrects the package check for metadata added by sci fetch; all scientific results are unchanged.

stereoelectroencephalography · early stopping · exact bounds · brain-computer interfaces · fixed-order optimization · binary64 arithmetic

Admitted 0 citing publications

References

6
  1. usesCertified feature packets cut bytes by 82.2% and kept all 160 fixed SEEG reader labels in eight STReEF patients under exact scoring sci:sha256:c000fc558baec2eed21eeaca82817f47a578b786cfaa41d43ec6f75c6af5ba48

    Its fixed-reader packet certificates informed the exact decision-agreement question. This study uses enforced pair score bounds and counts actual pair calculations. No code or data were copied from it.

  2. usesPartial requests kept all fixed phase commands and changed packet cost in seven SEEG rest recordings sci:sha256:0c3833998f471402e86b24e1fe226f662a0c827a6aabd6e519e01b31667ef449

    Its partial exact voltage requests for phase commands informed the related-method comparison. This study tests a fixed-order subset optimizer and lazy pair power calculations. No code or data were copied from it.

  3. supersedesExact bounds saved 17.7% of pair calculations and kept all full reader decisions in eight SEEG recordings with clipped pair scores sci:sha256:2ee5d1d6ff8e792fb698b6a21f10667ae638932ac2e41ebdf96e8ec7d60666a0

    This revision corrects the package hash checker and manifest builder so that they accept the metadata folder added by sci fetch. It adds a revision note and file comparison record. All source data, analysis code, proofs, numerical results, timing results, and figures are unchanged.

  4. usesOne shared gain constraint gave lower mean Brier error than eight pair gain constraints in eight reused SEEG records under a uniform gain model sci:sha256:c23a39e2554b405e4ae3bc346d7725666cbf215170a730abf032b586a85f3661

    This study uses its exact selected digital SEEG buffers, fixed unrestricted power readers, chronological splits, source metadata, and retrieval code. All 146 copied files match its manifest.

  5. usesdoi:10.18112/openneuro.ds004473.v1.0.2

    The study uses selected recordings and inherited reader inputs from OpenNeuro ds004473 version 1.0.2, recorded under CC0.

  6. doi:10.1088/1741-2552%2Facae0a

    This paper describes the movement-oscillation study cited by the dataset README and provides context for the public data; this study uses the archived dataset and recruits no new subjects.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.