Unverified50% confidenceCautionExact time
μ 和 σ 必须仅从训练集计算，若将验证集或测试集数据纳入 fit 会造成数据泄露
1
Sources
50%
Confidence
Long-term
Relevance
7/17/2026
First Seen
Sources
Related Claims
Verified若用含测试集的全量数据计算均值和标准差，会导致预处理泄漏，使评估指标虚高，无法真实反映模型泛化能力69% similarUnverifiedData leakage falls into two categories—target leakage and training set leakage—and causes models to perform well during testing but fail after deployment.67% similarUnverified预处理泄漏在小数据集上尤为显著，当训练集仅有数百张图像时加入测试集样本会明显改变均值和标准差估计67% similarUnverifiedThe true difficulty of model fine-tuning lies in the construction and cleaning of high-quality training data, preventing catastrophic forgetting, and inference deployment after quantization.66% similarVerifiedBenchmark data contamination refers to the phenomenon where, after a test set is made public, its questions may appear in the pre-training corpora of subsequent models, causing scores to reflect memorization rather than genuine reasoning ability.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/547581API
curl https://kongchang.com/api/v1/knowledge/claims/547581MCP
get_claim(id=547581)