Unverified50% confidenceFactExact time
MMD（Maximum Mean Discrepancy）是基于再生核希尔伯特空间的非参数分布距离度量，相比FID对小样本场景更友好、计算开销更低
1
Sources
50%
Confidence
Long-term
Relevance
7/15/2026
First Seen
Sources
Related Claims
Unverified量化测量域偏移程度的常用方法包括最大均值差异（MMD）和Fréchet Inception Distance（FID）73% similarUnverified可借助Fréchet Inception Distance（FID）或Maximum Mean Discrepancy（MMD）等分布距离度量来量化增强后训练分布与真实测试分布之间的偏移69% similarUnverifiedCWD 通过 Softmax 对每个通道的空间激活值归一化转化为概率分布，再用 KL 散度计算师生模型通道分布差异作为蒸馏损失67% similarUnverifiedMissing value handling requires judgment about the data missingness mechanism—Missing Completely at Random (MCAR), Missing at Random (MAR), or Missing Not at Random (MNAR)—as the optimal handling strategy differs under each mechanism.63% similarUnverified哈达玛变换可用于LLM超低比特量化的预处理，缓解离群值导致的精度崩溃问题63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/520685API
curl https://kongchang.com/api/v1/knowledge/claims/520685MCP
get_claim(id=520685)