Spaces:
Running
Running
File size: 1,445 Bytes
97afa54 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | <!doctype html>
<meta charset="utf-8">
<title>Reproduction: Semi-knockoffs</title>
<style>
body{font:16px/1.6 system-ui,sans-serif;max-width:52rem;margin:3rem auto;padding:0 1.25rem;background:#0f1116;color:#e6e8ee}
a{color:#7aa2f7}code{background:#1a1d26;padding:.15em .4em;border-radius:4px}
th,td{border:1px solid #2a2f3a;padding:.4rem .6rem;text-align:left}table{border-collapse:collapse;margin:1rem 0}
h1{font-size:1.5rem}.ok{color:#9ece6a}
</style>
<h1>Semi-knockoffs — reproduction</h1>
<p>arXiv:<a href="https://arxiv.org/abs/2601.23124">2601.23124</a>v1 ·
OpenReview <code>Xf9hJMGwDd</code> ·
authors' code <a href="https://github.com/AngelReyero/loss_based_KO">AngelReyero/loss_based_KO</a></p>
<table>
<tr><th>Claim</th><th>Subject</th><th>Headline</th></tr>
<tr><td>1</td><td>Valid p-values, no train–test split</td><td class="ok">type-I 0.017 at α=0.05</td></tr>
<tr><td>2</td><td>FDR control at level q</td><td class="ok">FDR 0.159 ± 0.024 ≤ 0.2</td></tr>
<tr><td>3</td><td>Null-feature optimisation stability</td><td class="ok">n<sup>−0.654</sup>, R²=0.996</td></tr>
<tr><td>4</td><td>Double robustness, compound rate</td><td class="ok">n<sup>−0.891</sup>, R²=0.981</td></tr>
<tr><td>5</td><td>Higher power than HRT + derandomisation</td><td class="ok">beats HRT 6/6 cells</td></tr>
<tr><td>6</td><td>Model-agnostic on Wisconsin Breast Cancer</td><td class="ok">type-I ≤ 0.067 across RF/NN/GB</td></tr>
</table>
|