Unverified50% confidenceBenchmarkExact time
On a dataset with 5,000 random features and random labels, if the top 20 features are first selected on the full data and then 5-fold cross-validation is performed, accuracy can reach 85% (87.7% averaged over 20 random seeds); after moving feature selection inside the Pipeline, accuracy falls back to 48% (50.0% average).
1
Sources
50%
Confidence
Long-term
Relevance
11/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn the benchmark, the random policy obtained an average of approximately 0.56 badges.76% similarUnverifiedOn purely random labels, the expected accuracy of any leak-free model is 50%, and being significantly above that baseline almost certainly indicates information leakage.76% similarUnverifiedUnder the condition of n=100 with an even split of positive and negative samples, the 95% confidence interval for 55% accuracy is roughly 45%–65%, which covers the 50% baseline, so the 5 percentage point gap may be random fluctuation.75% similarUnverified在5%显著性水平下测试100个随机因子，期望会有5个看起来显著（多重假设检验问题）75% similarUnverifiedOn the RadioML 2016.10a dataset, using 5 random seeds and a held-out test set of 44k frames, FuseAMC achieves 55.8% overall accuracy.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1142635API
curl https://kongchang.com/api/v1/knowledge/claims/1142635MCP
get_claim(id=1142635)